.8 :- v.rv| V. 41:51.... 1.: L 4. . .. . . . V . ,l 0:. . ‘ . ‘ ‘ un$aKllK| . . ,.i.t.(:.b . . V v . .. , ‘ . ‘ V . . . V . , ).ra. ‘ . ...Yll|(§3.. . avo. Lav'P..| ftll. ‘.v.| .\. 3|: .1....r> .’ u. . \f. AY'J 1.: r.‘:4 ~31 5. In: 4.3. .1 ‘3! V J. . . i I V ............i... u 1‘: .... ‘ , o ‘ ‘ . . . ‘ ‘ lid IAIII I' L ‘ L 21:... i I t . w. I ‘ ‘ .1. “ '. 1.. § . . .,_.I .. . . . a; 31.3 3% a? . .. 7 _ . . . w 13:: infu... g 3 2.9+? 53%;: 5&5 . ii. . . ERR”... n | I I'l I ‘ l - - SITYUB BRARISE IlllUlUllHlllllilHHillNlHUliWNW H lllli 31293 0089 This is to certify that the ' . dissertation entitled ‘ 4 :6 K ‘ C0144; M’Lwc’t” 092/6; Bjédu/ ’341’66/5214m'g .' ' amt am mm 4 gamma-17 Mari presented by fife W16 71 (1'4""((’]5 has been accepted towards fulfillment of the requirements for ill: A degree in givéjidi-M't EdLKQTL/e’m Maw Major professor Date 3‘919'6/73 MS U i: an Affirmative Action/Equal Opportunity Institution 0- 12771 w. . —._— _..-“ 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 .050 3,319“? _9_. . .. L .r ,4 -. ,ng 0' 9?};99 Jgtlootlége'? DEC ‘02 L205 MSU It An Affirmative Action/Equal Opportunity Institution 63‘:ch WG-9. I a. u I l - S...“ SELF-CONFIDENCE AND BASEBALL PERFORMANCE: A CAUSAL EXAMINATION OF SELF-EFFICACY THEORY BY Thomas Robert George A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Physical Education and Exercise Science 1993 (I) ‘ A 0,. 7‘“ oc- s v a .h A“. '0. V- I o ‘ “A. -‘ H favu u. ‘ A .‘ . r.- A . .b‘ub-. FFA “it: ABSTRACT SELF-CONFIDENCE AND BASEBALL PERFORMANCE: A CAUSAL EXAMINATION OF SELF-EFFICACY THEORY by Thomas Robert George The relationship between self—confidence and performance has been 9' examined by sport scientists for years. Yet, this relationship is still not fully understood. Though many theories of self-confidence have been proposed, self—efficacy theory currently provides the strongest framework from which to investigate self-confidence in sport. This study employed path analytic techniques in an effort to examine the causal relationships in Bandura’s model of self-efficacy in a field setting. Male intercollegiate and high school baseball players (N=53) completed self—report measures over a nine- game period during their respective seasons. Perceptions of self-efficacy (confidence in hitting performance). competitive state anxiety, effort expenditure, and performance were assessed, as well as an objective measure of performance (contact percentage). It was hypothesized that previous performance and anxiety would significantly predict self-efficacy beliefs, and that self—efficacy would mediate the effects of previous performance and anxiety on effort expenditure and hitting performance. Moderate support for Bandura's model was found in that higher contact percentages were predictive of stronger percepts of efficacy in five games, and lower levels of somatic and cognitive anxiety were associated with stronger self-efficacy beliefs in seven gates. Htv>‘l" u.v'¢“‘ games. In turn, self-efficacy was a predictor of effort and hitting performance in six of the nine games. In all cases, higher percepts of efficacy were associated with increased effort expenditure and greater hitting performances. Results are discussed in relation to the ecological validity of previous causal examinations of self-efficacy theory, as well as the utility of self-efficacy theory as a framework for investigating the self—confidence-performance relationship. ACKNOWLEDGEMENTS I would like to express my sincere appreciation to Dr. Deborah L. Feltz for all the encouragement and insight she offered not only during the dissertation process, but throughout my doctoral education. I would also like to thank the members of my committee, Dr. Mary Ewing, Dr. Vern Seefeldt, and Dr. Galen Bodenhausen for their valuable suggestions and critical input. The time and effort put forth by my committee members certainly improved the Quality of this work. A special thank-you is also extended to Dr. Feltz, Dr. Ewing, and Dr. Seefeldt for all of the professional advice and guidance they offered during my graduate training. I would also like to thank my fellow graduate students for their camaraderie during my years as a doctoral student. Finally, I would like to thank my wife, Kate, for all of the encouragement and support she has given me throughout this endeavor. iv av” V" CUuL-r v O ’1‘. v I . .n‘r . C. r n 'fil' r i V'u r41 TABLE OF CONTENTS List of Tables ...................................... vii List of Figures ..................................... viii Chapter I: INTRODUCTION Nature of the Problem .......................... 1 Statement of the Problem ....................... 13 Hypotheses ..................................... 13 Delimitations .................................. 14 Definitions .................................... 14 Hitting Self-efficacy ..................... 14 Objective Hitting Performance ............. 14 Perceived Competitive Anxiety ............. 15 Perceived Effort Expenditure .............. 15 Subjective Hitting Performance ............ 15 Basic Assumptions .............................. 15 Limitations .................................... 15 Chapter II: REVIEW OF LITERATURE Overview of Self—efficacy Theory ............... 16 Self-efficacy Research in Motor Performance .... 18 Performance Accomplishments ............... 18 Vicarious Experience ...................... 21 Verbal Persuasion ......................... 23 Emotional Arousal ......................... 26 Causal Examinations of Self-efficacy ........... 29 Chapter III: METHOD Subjects and Design ............................ 36 Dependent Measures ............................. 36 Hitting Self-efficacy ..................... 36 Competitive Anxiety ....................... 37 Objective Hitting Performance ............. 38 Subjective Hitting Performance ............ 38 Effort .................................... 38 Procedure ...................................... 39 Treatment of the Data .......................... 40 Chapter IV: RESULTS Preliminary Analyses ........................... 43 Hitting Performance ....................... 43 Self-efficacy ............................. 46 Anxiety ................................... 49 Effort .................................... 53 Correlations Among Independent Variables ....... 55 Group Differences .............................. 56 Path Analysis .................................. 59 Summary of Path Analysis ....................... 72 Chapter V: Preliminary Results Path Analysis Conclusions Practical Implications Limitations of the Study Future Directions Appendix Appendix Appendix Appendix Appendix Appendix Appendix Appendix Appendix Appendix Appendix List of References A: B: J: K: DISCUSSION Hitting Self-efficacy Scale Competitive Anxiety Scale .............. Subjective Hitting Performance Scale, Perceived Effort Scale Consent Form ........................... Means and Standard Deviations for all variables ...................... Intracorrelations for Independent Variables .............................. Intercorrelations Among Independent Variables .............................. Path Coefficients for the Hypothesized and Full Models ........... Means and Standard Deviations for Self-efficacy Questionnaire Items Data Coding Sheet Raw Data vi 10. ll. 12. 14. 15. 16. LIST OF TABLES Means and Standard Deviations of All Variables by Group and Wave Intracorrelations Intracorrelations Intracorrelations Intracorrelations Intracorrelations of of of of of Hitting Performance ........ Self-efficacy .............. Cognitive Anxiety .......... Somatic Anxiety ............. Effort ..................... Correlations Among Independent Variables by Wave Correlations for Hypothesized Paths by Group and Wave Path Path Path Path Path Path Coefficients Coefficients Coefficients Coefficients Coefficients Coefficients Model by Wave Path by Wave Coefficients for for for Hypothesis 1 .............. Hypothesis 2 .............. Hypothesis 3 .............. Hypothesis 4 .............. Hypothesis 5 .............. the Hypothesized Means and Standard Deviations for Self-efficacy Questionnaire Items by Group and Wave ........ vii .Ao tr» LIST OF FIGURES Bandura’s Model of Self-efficacy .................. Conceptual Model of Hypothesized Relationships Among Variables ................................ Hitting Performance of College and High School Players by Wave ............................... Subjective Performance of College and High School Players by Wave ............................... Self-efficacy of College and High School Players by Wave ............................... Cognitive Anxiety of College and High School Players by Wave ............................... Somatic Anxiety of College and High School Players by Wave ............................... Perceived Effort Expenditure of College and High School Players by Wave ................... viii CHAPTER I Introduction W For years, athletes and coaches have espoused the importance of self-confidence as an essential component of successful athletic performance. Typically, coaches tell athletes to “think like a winner“ or “believe in yourself" in an effort to boost athletes' self-confidence. Similarly, when athletes are performing well they are said to be "very self-confident": whereas, they are often characterized as having "lost their self-confidence“ during poor performances. Self-confidence has received so much attention in the world of sport that it is one of the most frequently cited psychological factors thought to affect athletic performance, and may be the most critical cognitive factor in sport (Feltz, 1984; Gill, 1986). Researchers in the sport sciences have devoted a considerable amount of attention to the concept of self-confidence and the relationship between self- confidence and athletic performance. Generally, the research has provided support for the notion that one's level of self- confidence is related to one's athletic performance. For example, one of the most consistent differences between successful and less successful elite athletes is that successful elite athletes report greater self-confidence (Gould, Weiss, & Weinberg, 1981; Highlen & Bennett, 1979, 1983; Mahoney & Avener, 1977; Meyers, Cooke, Cullin, & Liles, 1979). .‘l . rynfl‘ 7‘9" OiudrbAdv. up PC" “to- ~v‘ ; Q AA . ‘ ~~ " "V&-eo~ “:VSAfi 1"“ UA.€ 4 ‘4 O F‘ " nae J 2 vThe construct of “self-confidence“ has been defined in a number of ways in the research literature. Concepts such as ”performance expectancies" (Dweck, 1978), "perceived ability" (Nicholls, 1980), and “perceived competence“ (Harter, 1978) have all been used in the achievement and motivation literature to describe the perception of one's ability to successfully perform a given task. In the sport and motor performance literature, self-confidence has been conceptualized in similar terms to those described above as well as in other ways. Griffin and Keogh (1982) have coined the term "movement confidence" to describe an individual's feeling of adequacy in a movement situation as both a personal consequence and a mediator in that situation. Vealey (1986) developed a sport-specific model of self- confidence in which “sport-confidence“ was defined as the belief in one's ability to be successful in sport. Perhaps the most extensively used theory for examining self- confidence in sport and motor performance has been Bandura's (1977, 1986) self-efficacy theory (Feltz, 1988a). Bandura (1986) defines self-efficacy as “people's judgments of their capabilities to organize and execute courses of action required to attain designated types of performances" (p.391). The construct of self-efficacy is more concerned with one's judgment of performance potential given one's skills, rather than the sheer number of skills one possesses. Thus, whereas self-confidence is generally perceived as a more global trait, self-efficacy is a situation—specific form of self- confidence in which individuals believe that they can do whatever needs to be done in a specific situation. Self-efficacy theory purports that when the necessary skills a I..- E"'C5~71' n" nary" xiv: . ry. U ,- erg—ear... ‘vbvpuy; v a ‘ AMA i to U....LI‘ quake. Ar: CAnV wovtbv ' p nfitAfwfi. -oo-d-u-a5 - - I‘V‘fiHY -C‘A“- a p “EY‘Ay-w- 3 - ~v-u.: scarce c 3 skills and appropriate incentives are present, one's self- efficacy will predict actual performance. Moreover, self- percepts of efficacy determine the choice of activities attempted, the amount of effort expended and the persistence to complete the activity, as well as thought patterns and emotional reactions during actual and anticipated encounters with the environment (Bandura, 1986). In turn, self-percepts of efficacy are based on four principle sources of information: performance accomplishments, vicarious experiences, verbal persuasion, and emotional arousal. Bandura (1986) asserts that "enactive attainment” or performance accomplishments provide the most influential source of efficacy information because they are based on actual mastery experiences. Vicarious experiences provide efficacy information through social comparison processes. Efficacy information from verbal persuasion may include social persuasion, self-talk, imagery, and other cognitive strategies. Emotional arousal can affect perceived self- efficacy through cognitive appraisal such as associating physiological arousal with fear and self-doubt. According to Bandura (1977), self-efficacy operates as a common cognitive mechanism for mediating the effects of these sources of information on performance. Figure 1 illustrates the relationship between the sources of efficacy information, self-efficacy, and athletic performance. A number of studies in the sport and motor performance literature have provided support for Bandura's (1977) predictions of the effects of various sources of efficacy information on perceived efficacy and performance. Feltz, Landers, and Raeder (1979) found that subjects in a .Ammma .Noaom Eouuv Shoonu m.musocom SD pmuofipmua mo mnuouuno uneconuxwofl>n£on one .mcofiumuoooxm xomofluum .nofiuofiu0mcfi xoooflmmm wo mwou50m Donne ommzuon magmnoflumHmm .H musmfim 825953 Eo>> £80 wzmwtka 50301.9 682$de term 8650 mO_>cn_ commameoa _m€o> moocotoqu , 30:85 m fl 9coE:w__aE8o¢ oocwESton. an“ e a..- €"cr* - -J- .u , . pr.fiy-‘ :y- -U-oht‘ recan vvor Vs I \—\ .1:- ~ “ac-Ia ‘ \ 1' al' . 1' n p Q QC Ce .e A t . . . y ._ V. n: B . QC .4 a; a k at . . u x 1 E E E. fin S A: e nu. A: . C e u .. 1 . .- .l h. C . : .K a; .3 .: «.5 Au 3 a; a; Q r r e c .. . £ . nu a 4 CV .mu r r C g a» r we EL 5 4. p“ t. a. «a. p. .... a» .t Lu 3 no my .9. ”A .l S ~\u 11‘ s b.» “be .u.‘ s u A» A.» o. 5 participant modeling group outperformed and exhibited stronger expectations of self-efficacy than did subjects in a live and videotape model groups. McAuley (1985) found similar results using a high-avoidance gymnastics skill. Other research has focused on the effects of modeling (vicarious experiences) on self-efficacy and performance. Gould and Weiss (1981) found that viewing a similar model perform a muscular endurance task resulted in better performance and higher ratings of perceived self-efficacy for performing that same task. Similar findings have been reported by McAuley (1985), and McCullagh (1987). Verbal persuasion as another source of efficacy information has received little attention in the sport and motor performance literature. However, Peltz and Riessinger (1990), Ness and Patton (1979), Weinberg, Gould, and Jackson (1979), and Wilkes and Summers (1984) have provided support for the influential effects of persuasion on self—efficacy. Several studies have also investigated the effects arousal asserts on self-percepts of efficacy (Feltz, 1982, 1988b: Feltz & Mugno, 1983; McAuley, 1985: Lan & Gill, 1984). The results of these studies tend to show that arousal influences performance through its affects on perceived self-efficacy. Based on the preceding research in the sport and motor performance literature, self-efficacy has been shown to be influenced by the four factors postulated by Bandura's (1977) theory. In turn, self-efficacy has been shown to be correlated with athletic performance. However, the correlational designs of most of the preceding studies have not permitted inferences to be made with regard to causality and direction of the relationship. In an effort to ascertain r‘P‘GV -509- an? 1 . . AH"; ‘ " v‘oeefii :r “.0 va‘ ’ -E q a . ave-hr? v- ~‘v.c a U . \— “. '“F‘vq ”V‘O'-a. VI. coe‘.‘g PErfor 3" . . clnclh A4 ‘ pm 5 VS ‘ “L t) .1 ‘M Ah,”- 5 Va. 6 whether a causal relationship existed and to determine the direction of such a relationship, Feltz (1982) examined the predictions of Bandura's self-efficacy model, as well as those of an anxiety based-model, using path-analytic techniques. Subjects were measured across four trials on self-efficacy, heart rate, self-reported anxiety, and approach/ avoidance of a back dive. The results of Feltz's study indicated that self- efficacy was a significant predictor of back—diving performance. In addition, a reciprocal cause and effect relationship between self-efficacy and diving performance was found, providing support for Bandura‘s model. However, contrary to Bandura's model, performance exerted a greater influence on self-efficacy than self-efficacy exerted on performance. Moreover, self-efficacy was not the strongest predictor of performance, as Bandura would assert. Previous performance was also found to have a direct effect on future performance after the first performance trial. These findings prompted Feltz to propose a respecified model composed of previous performance and self-efficacy as predictors of motor performance. Subsequent goodness-of—fit analyses revealed that the respecified model explained more variance than did Bandura's model, and thus proved to better fit the data. Feltz and Mugno (1983) replicated the 1982 study and found similar results. The respecified model accounted for more variance and fit the data better than did Bandura's self-efficacy model or an anxiety-based model. Further support for the influences of past performances and self- efficacy on future performance has been provided by McAuley .. .1 €5L-CI 1’ «Au 9v y CL f» V. . ¢ e. V. HI: I I... r\ .c 1. ‘t In :1 . . an». r. nu. .. . . a. my. h. av U .1 In (C O .QC .1. O nu. Q» :s 3 . . H a . e .2. a C. .3 C .3 .1 £1 .fl e W .C a. e e v . S e . e . . C. 1. e a.» d v . rt inn 9 S e u“ an“ .ni. .3 CC .C u u a: .. a .3 Kb. v . an a» QC A o t. 1. .Q «h v a >e . . I. .C a.“ .n~ n. v. S u: S “u. 4. no. .u~ Ci. .9. Lu .5 .0. (1985) and Feltz (1988b). Feltz (1988b) noted that previous research examining the causal relationships in Bandura's model has employed female subjects only. Thus, it was uncertain as to whether the same results would be found with male subjects. In an effort to examine possible gender differences, Feltz (1988b) contrasted the path analysis models of female and male subjects using the respecified model of Bandura's (1977) self-efficacy theory. The same high-avoidance diving task used in her earlier work (Feltz, 1982: Feltz & Mugno, 1983) was employed. Results indicated that the respecified model fit the data better for females than for males, though much variance was unexplained in both groups. Overall, however, the 'relationships in the respecified model tended to be supported. Taken together, the previous research indicates that Bandura's (1977) model of self-efficacy, and more specifically Feltz's respecified model, has some utility for predicting sport and motor performance. These studies indicate that self-efficacy is a strong predictor of motor performance. Moreover, a causal and reciprocal relationship between self-efficacy and performance has been demonstrated. However, contrary to prediction, self-efficacy has not been the only predictor of performance. Rather, as Feltz (1982) has proposed in her respecified model, previous performance as well as self-efficacy have been shown to be strong predictors of subsequent motor performance. Although the eXtant sport psychology literature has tended to support Bandura's (1977) theory of self-efficacy, the research is not without limitations. First, in the four . q - 0'1 . F ‘ " i “-v‘AO“ :CJnar- e-bea‘v- A C a. p‘tr-or.t. 8 studies using path analytic techniques to examine Bandura's model, very few performance trials were allowed over a short period of time. Subjects generally performed a few trials within a one hour time period (or less). Though self- efficacy was found to be a significant predictor of performance, past performance was an even stronger predictor. As Feltz (1988a) noted, perhaps subjects' self-percepts of efficacy were only partially influenced due to the limited number of performance trials and/or the short time period in which they performed. Furthermore, subjects in these studies had little experience with the task being performed. Thus, efficacy expectations were likely based on experiences perceived to be similar to the experimental task, but may or may not have been relevant to the performance of the task. The present study employed subjects who had extensive experience with the hitting task, and thus had a strong sense of personal efficacy with regards to their ability to perform the task. Second, the existing self-efficacy research in the sport and motor performance literature has generally examined non— athletic subjects performing novel tasks under controlled, invariant conditions. Though this research has generally supported the basic tenets of self—efficacy theory, the extent to which self-efficacy predicts motor performance under changing, dynamic conditions remains unclear. For example, is the relationship between self-efficacy and performance dependent upon the type of task being performed or the conditions under which it is performed? In an effort to address this issue, the present study examined subjects' performance of an open skill (hitting a baseball) under ('4 \ i Av- ha. C Pain up-» "I“? .5“ ‘g; Inc-J Vs.“ AV... \J‘ on V I ,7 1.1 a~v~ ‘5‘. he . 1e e a wee e 1e t 1 to te nd 5"». . o k u on. \r s n. u..- pcr C . . 'F-i‘ .‘hli- A ‘ e::;ngs. a o '5’.” avg. . . db..- (aria F.1r0' (I"' V qflv“ U a \ v '9 'I ca cg ya ». a: at we be 9 variable environmental and situational conditions. Moreover, with the exception of a few studies (e.g. Gayton, Matthews, & Burchstead, 1986: Lee, 1982; McAuley & Gill, 1983), the research has examined the self— efficacy/performance relationship in "artificial" sport settings. That is, the research has not examined athletic performance in actual competitive situations. Though this is not a weakness of the previous research per se, it is a limiting factor if Bandura's (1977) model of self-efficacy is to be used to explain and/or predict athletic performance. Before such a generalization can be made, research examining athletic performance in actual competitive situations is needed in order to delineate the causal and directional relationships of self-efficacy theory. The present study attempted to address this concern. Third, the previous research has generally employed objective measures of performance in order to predict future performance and self-efficacy. As Bandura (1986) has noted, individuals are more influenced by how they perceive their performance successes and failures than by their performance attainments per se. For example, two baseball players may experience the identical hitting performance (e.g. one hit in three at-bats) differently. Thus, objective performance measures may not accurately reflect one's subjective appraisal of past performances. It may, therefore, be more important to assess how one perceives his or her past performances in order to assess the effect of past performances on self-efficacy expectations. Furthermore, as Feltz (1988a) has nOted, previous research has measured only performance and not other measures of behavior such as ‘:AY" €5~V‘ x 3‘53; . in?“ 3" A" r- e“"" 1.9:?" e IMO. . Y.A V‘~‘v “A v.“ O a: re C .. u--er V V“)- 'v A “g . A H7 ~ 3 1' I Q “~d ‘ lO effort. The present study employed both objective and subjective measures of performance, as well as a measure of effort as dependent variables. Finally, the previous research (e.g. Feltz, 1982; Feltz & Mugno, 1983; McAuley, 1985) has used female subjects predominantly to examine Bandura's (1977) self-efficacy theory. Prior research indicates that males and females differ in perceptions of self-confidence in relation to a variety of performance situations ( Lirgg, 1991: Maccoby & Jacklin, 1974). Likewise, perceptions of anxiety and arousal have been shown to differ among male and female subjects (Borkovec, 1976). These differences may affect the proposed relationships in Bandura's model. Therefore, this study assessed anxiety in order to determine whether males' perceptions of anxiety influenced self—efficacy and/or performance. In summary, the previous research has provided support for Bandura's (1977) self-efficacy theory. Self-efficacy has been shown to be a strong predictor of motor performance. Moreover, a reciprocal relationship between self-efficacy and performance has been demonstrated in a number of studies. However, the data do not fully support Bandura's predictions. Feltz and associates (Feltz 1982, 1988b; Feltz & Mugno, 1983) have consistently found that after the first performance trial, previous performance is as strong a predictor of future performance as self-efficacy. Bandura's (1977) model asserts that the effects of previous performance on future performance are mediated by self-efficacy, and therefore, self-efficacy should be the strongest predictor of performance. Bandura and his colleagues have found support P “ [r g Fq:p “ .r- t'“"’" z, cvcv~~ VCD" 7" ’( J (D ’ f I h 11 for such a relationship in their research efforts (Bandura & Adams, 1977; Bandura, Adams, Hardy, & Howells, 1980; Bandura, Reese, & Adams, 1982). As previously noted, this has not necessarily been the case in the sport and motor performance literature. Perhaps these inconsistencies are due to some of the aforementioned research designs. The temporal aspects of performance, the types of tasks being performed, the conditions under which tasks are performed, or the number of past performances may limit the influence past performance exerts on self-efficacy, and therefore limit the ability to predict future performance from measures of self-efficacy. Furthermore, objective measures of performance may not be tapping the influence performance exerts on self-percepts of efficacy, and vice versa. The key to understanding the relationship between self-confidence (self-efficacy) and sport performance may lie in the longitudinal examination of athletic performance. By measuring self-efficacy and performance in an actual sport setting over a period of time, the predicted relationships posed by Bandura's (1977) self- efficacy theory may become more evident. Thus, a conceptual model was proposed to test the network of path analytic relationships among anxiety, self-efficacy, effort, and performance across games (see Figure 2). Though only the first three waves of the model are shown, the network of relationships were hypothesized to persist across nine waves. Based on previous research (Bandura, 1977; Feltz, 1982; Feltz & Mugno, 1983), anxiety as well as past related performance accomplishments is expected to exert direct effects on self—efficacy. Self-efficacy, in turn, effects .1 W.IV(? 3 We] v e) WriVl-P .moEmm mmouom mocmau0muma poo .uu0mmo .homoflwmoumaom .xuoflxcc mcoEo moflcmoofluoaon mo xHOSDmG on» mo Hobos amouomocou .m muamflm 325.5 \ 1 , _ m 0:80 m 3853.8 - No.50 Alli—m 3853.3 I 7250 _‘ul.|l.— ~>uaoEo..=om_ — oogEotom - 2 mrEEM - , v5 m _...._.._._ . .. E i E E m m>oz - N m>mz . H o>mz ..' I?! arm-1(be”: 15"”2‘ ‘ b- v»- ‘5”fah ”wk-1“}, . " “OYCAY‘ 2'!“ 5Vbl 13 performance directly and indirectly through effort. Finally, performance has been shown to directly influence subsequent anxiety as well as self-efficacy (Feltz, 1982). W The major purpose of this field study was to examine the causal and directional relationships predicted in Bandura's (1977) self-efficacy theory and found in previous path models using male athletes in actual sport performance situations. More specifically, self-perception measures including perceived self-efficacy (confidence in hitting performance), competitive state anxiety, effort expenditure, and subjective performance ratings, as well as objective performance ratings, were obtained from collegiate and high school baseball players over a portion of a baseball season. These variables were used to test the predictions of a conceptual model based on Bandura's (1977) theory and previous path models of the self-efficacy-performance relationship. This model was tested against a fully recursive model that contained all of the pathways among the variables. Hypotheses Based on Bandura's (1977) theory of self-efficacy and the extant literature, the following hypotheses were set forth for each wave: 1. Performance will be a significant predictor of anxiety in the proposed model. Furthermore, performance will be negatively related to anxiety. 2. Anxiety and past performance will be significant predictors of self—efficacy. Anxiety will be negatively related t0 picsiti‘u’el'r’ 3. Se .‘q vine pro; .u V‘L {.IESICC se- 1".“ u . . "‘fi§ ~O-V A n ’1 m‘ 1.155 ,1 A V ‘ cCZIE-ECES pa TI- acneraliza: activizies Cm0t be m I an" h.‘ . “C ‘ M ”x Terms 533‘! are l 14 related to self-efficacy, whereas past performance will be positively related. 3. Self-efficacy will be the only predictor of effort in the proposed model. 4. Perceived self-efficacy and effort will be the only predictors of performance in the proposed model, with self- efficacy being the strongest predictor. 5. Subjective performance ratings will more strongly predict self-efficacy than will objective performance ratings. 1' . . This Study is limited to collegiate and high school male athletes participating in the sport of baseball. Generalizations to other levels of sport participation, to activities other than baseball, or to female populations cannot be made. E' . . Terms and operational definitions which apply to this study are listed below.' Hi;§ing_aelf;effigagy_- an individual's judgment of his capabilities to organize and execute courses of action to attain designated performances; for this study, strength of hitting efficacy relating to a subject's ability to hit a pitched baseball within fair territory was assessed before each game. WW- hitting performance measured by contact percentage, a representation of the number of times the ball was hit within fair territory ‘ . {9-3: Ive P;- F' VF'V w ”A’Weti: i‘ yup-M . version o 15 relative to number of at-bats. .Eerceiyed_semnetitiye_anzietx_- measured by the Competitive State Anxiety Inventory - 2 (CSAI-Z), a shortened version of the 27—item self-report state anxiety inventory developed by Martens, Burton, Vealey, Bump and Smith (1983) which measured how a subject felt just prior to hitting. EgIQeiyed_effgrt_expenditu1e — a self—report measure which required subjects to indicate on a 10-point Likert scale how much effort they exerted while hitting. snhiectiMe_haseball_nerfermance_ratins_- a self-report measure which required subjects to rate their performances on a 10-point Likert-type scale; subjects completed a performance rating after each game. . . 1. Athletes possessed the requisite abilities and were motivated to perform baseball skills. 2. Athletes answered the self-report measures honestly. I. . . 1. External factors (factors not measured in the study) that may have influenced self-efficacy and/or performance were not controlled. L——___-. p “IVR‘H A. ' terlt \Jb . enbnr n .v’de _ Jesus ‘VVOOVO‘ ~- A; I- “ "n‘n: 8.x? J. D D‘Avu e::;:a':; met: V :‘f:".‘“ F e y; ~06.“er VOA ‘ \ ~A'“ ‘ F." aw. vying v ‘Ov a“ a ‘ e A u- 2.“ hit-:1. . b-‘¥.\ilb~.b-~ I ~ ‘1 '5 V’\FA.“ ._ wan. “Out. I v . NON” \ \ ~w.es “ave . Nay ‘ i Mess .ne .~ “\A A..‘ - ~y~~ifi n ‘ p 5““. ‘éeevb 5.:- “PVC“ Vu- :‘tn‘sd‘ulanc‘ W 4-5 Y‘fiy ~ ~ 1' 01‘ c. ‘1; CHAPTER II Review of Literature Much of the research investigating self—confidence and motor performance has employed Bandura's (1977,1986) theory of self—efficacy. Generally, researchers have found self- efficacy theory to be a useful framework from which to examine the relationship between one’s sense of physical competence and actual sport or motor performance. The extant research has examined the self—efficacy/performance relationship in a number of ways, including correlational and causal assessments. Typically, support has been found for the propositions in Bandura's model, though a number of studies have reported contrary findings. In an effort to address the utility of self-efficacy theory, as well as compare and contrast the research in this area, this chapter examines the self-efficacy research conducted in the motor performance area. First, a brief overview of self—efficacy is provided. Second, the relevant literature relating to the four sources of efficacy information is addressed. Finally, research examining the causal relationships in Bandura's model of self-efficacy is reviewed. : . ESJE-EE' 1 Self-efficacy theory, developed within the framework of a social cognitive theory (Bandura, 1986) poses self-efficacy as a common cognitive mechanism for mediating people's motivation, thought patterns and behavior. As noted in 16 Chapter 1, S=~ I ‘ “A “ ' v-‘ir f. h .4 ”1:43. baya ‘ u «an; n’ ‘ y-vUO ‘ h e's let": . ‘ . than itfifl’v' 2 I'- I' F i. t Atuvovu' .uv‘ --.:. ‘cvaqmn \- , ivOO ‘5‘» U‘ . area r 1n;- nb :oir . Luv “.1 ”Hit. 1 n ': ::¥:¥Oe s be-.. CC certam Cf}; .-.m;nv ‘.“ . I. . I‘- Udufiic ’ g, v -e ~n- '1 pa “A” no..- My gut». -A- I n 3" ."‘C an \- ‘bQAA‘, “u “A . an A" ' rd: Y‘v HUIuy‘v‘. . Prccessizg < i.‘ . '1-“N ‘ ~“‘“Lra' 7 us . Wye. 'ir‘q it ‘ . F A . iv- ‘V‘u‘aque “‘v . - ::‘slaslcr‘el 3““ Mfi‘ ' . "VC.L.yllsr: A. m2"; may” w“ Iv,“ . “it'd“ . a» v“ e. 5 & ;r3CeSS W: 17 Chapter 1, self-efficacy refers to people's judgments of their capability to perform at given levels, and determines people's levels of motivation as reflected in the challenges they undertake, the effort they expend in the activity, and their perseverance in the face of difficulties. However, self-efficacy is a major determinant of behavior only when proper incentives and the necessary skills are present. People's self-efficacy beliefs are also reciprocally related to certain thought patterns that influence motivation (Bandura, 1986). For example, self-efficacy beliefs influence people's success/failure imagery, worry, goal setting, and attributions (Feltz, 1992). Expectations of self-efficacy are, in turn, a product of a complex process of self-persuasion that relies on cognitive processing of diverse sources of efficacy information (Bandura, 1990). These sources of information include performance accomplishments, vicarious experiences, verbal persuasion, and physiological states. Performance accomplishments provide efficacy information through one's own mastery experiences, whereas vicarious experiences provide efficacy information through a social comparison process with others. Persuasory information includes verbal persuasion from significant others, as well as self-talk, imagery and other cognitive strategies. One's physiological state or condition can also provide efficacy information through cognitive appraisal such as associating physiological arousal with fear and self-doubt. Much of the research investigating self-efficacy in the motor performance literature has focused on one or more of the sources of efficacy information. The next section addresses the motor 18 performance research by organizing it around the four sources of efficacy information. Self-Efficacy Research in Motor Eerfgrmangg W Performance accomplishments provide the most dependable and powerful source of efficacy information because they are based on one's own mastery experiences (Bandura, 1977, 1986). Repeated successful performances lead to heightened percepts of efficacy, whereas consistent failure causes one to lower efficacy appraisals. Likewise, self-efficacy influences performance (Bandura, 1986). Heightened percepts of self- efficacy will lead to superior performance, whereas low efficacy expectations will lead to poor performance. Thus a reciprocal relationship exists between performance and self- efficacy in which previous performance will influence one's efficacy expectations, which in turn will affect future performance. The influence performance exerts on self-efficacy may be mediated by a number of factors including the amount of effort expended, the situational characteristics of the performance, the patterning of success and failures, and the amount of external help received (Bandura, 1986). For example, succeeding on a simple or easy task provides little information concerning one's abilities, and thus may not alter efficacy expectations. However, succeeding on a difficult task provides valuable information about one’s competencies and therefore should raise self-percepts of efficacy. ,mfflt ’fiib:'v\ o A'P ‘9 .n 3. v L. In ""n r{" . yAbdauw - p H oAr R .v. U "‘Vfl" Db. UtA i . nun: A \ du‘v C a . QAFI‘A .— ”.54.- ‘ Ann-fr 'u~-,,‘ ., NV-vn .- ‘svvl “Sr-t . e '. (Y) 19 (Research in the motor performance literature examining the effects of performance accomplishments on efficacy expectations has generally found that percepts of efficacy are influenced by previous performance attainments. For example, Feltz, Landers, and Raeder (1979) provided support for Bandura's contentions that mastery attempts provide the strongest source of efficacy information. In this study, subjects were assigned to a participant modeling, live modeling, or video-taped modeling condition in which they observed a model perform a high avoidance motor task (back dive). After viewing the model, the subjects attempted the same task. Results indicated that subjects in the participant modeling group, in which they were physically guided through the task, reported higher levels of self- efficacy and produced greater back-diving performance than subjects in either of the other two groups. McAuley (1985) conducted a similar experiment using a basic, but high-avoidance gymnastic stunt. Again, subjects were assigned to one of three conditions: aided participant modeling in which the subjects received verbal and visual instruction, as well as physical guidance while practicing the stunt; unaided participant modeling, which was similar to aided participant modeling except that no physical guidance was provided; and a control condition in which subjects viewed an irrelevant gymnastics videotape. Results revealed that both modeling groups reported significantly higher self-efficacy and performance scores. More specifically, aided participant modeling subjects performed significantly better than unaided participant modeling subjects, but the two groups did not differ on efficacy 20 expectations. These findings also support Bandura's- assertions regarding the powerful influential effects of performance accomplishments on percepts of self-efficacy. The influence of performance accomplishments on self— efficacy has also been examined in terms of perceived psychological momentum. Shaw, Dzewaltowski, and McElroy (1992) attempted to delineate the causal dimensions of psychological momentum by examining self—efficacy as a potential mediator of the relationship between performance and perceived psychological momentum. In this investigation, subjects were assigned to either a ”success" or ”failure" condition in a free throw shooting task. Results indicated that successful performances led to stronger efficacy beliefs and were associated with heightened perceptions of psychological momentum. However, changes over time in perceived momentum were not associated with changes in self— efficacy beliefs. Performance failures were associated with decreases in perceived psychological momentum, but did not influence percepts of self-efficacy. However, subjects who experienced failure exhibited a significant relationship over time between self-efficacy and momentum. Decreases in self- efficacy were associated with decreases in perceived momentum. This study provides support for previous research in that successful past performance was found to influence self-efficacy beliefs. In addition, partial support was found for the proposition that self-efficacy mediates the influence of performance on perceived psychological momentum. However, more research is needed before the role of self- efficacy in the psychological momentum process is more clearly understood. 21 . . . Research has shown that modeling facilitates the performance of a variety of motor skills (Carroll & Bandura, 1985; Feltz, 1982: Martens, Burwitz, & Zuckerman, 1976; McCullagh, 1986; 1987), as well as enhances self-percepts of efficacy for performance (Feltz et al., 1979; George, Feltz, & Chase, 1992; Lirgg & Feltz, 1991; McAuley, 1985). According to social cognitive theory, modeling influences operate primarily through their informative function, and are governed by four component processes: attention, retention, production, and motivation (Bandura, 1986). Attentional and retentional subprocesses are hypothesized to influence the acquisition of skills, and production and motivational subprocesses are hypothesized to influence performance. Efficacy information obtained through observing others engaging in a task is an important source of information when observers themselves have never performed the task. Live and filmed modeling have been shown to increase perceived efficacy in motor performance tasks (Feltz et al., 1979; Gould & Weiss, 1981; McAuley, 1985). The effectiveness of modeling on one's efficacy perceptions and performance depends, however, on a number of modeling variables (Bandura, 1986). For example, research has shown that model status and model competence mediate the influential effects of modeling (Bandura, 1969; Baron, 1970). Lirgg and Feltz (1991) found that observers of a skilled model, regardless of the model's status, exhibited higher efficacy expectations and superior performance relative to subjects who observed an unskilled model. McCullagh (1986) found that subjects who viewed a high status model performed better than subjects who viewed a 22 low status model. Similarities to the model in terms of performance or personal characteristics have also been shown to enhance the effectiveness of modeling procedures on subjects' efficacy perceptions and performance (Brown & Inouye, 1978; George et al., 1992; Gould & Weiss, 1981). For example, Brown and Inouye (1978) found that similar models were more effective in influencing observers’ feelings of learned helplessness than were dissimilar models. Gould and Weiss (1981) examined the effects of model similarity and model talk on muscular endurance and self-efficacy. The model-similarity results revealed that similar model subjects extended their legs significantly longer than dissimilar model and control subjects. In addition, similar-model subjects had higher and stronger self—efficacy beliefs than dissimilar-model subjects. It may be reasoned that the demonstrator-observer similarity increases the personal relevance of vicariously derived information for the observer, thereby enhancing or detracting the observer's self-efficacy perceptions and performance. Unfortunately, the design employed by Gould and Weiss (1981) did not permit the determination of the most salient dimensions of model similarity. More specifically, the authors used a nonathletic female as a similar model, and a male varsity athlete as a dissimilar model. Although the subjects (all nonathletic females) perceived the female nonathlete as being more similar and the male athlete as being less similar, it remained unclear as to whether this was due to similarities/dissimilarities in sex or ability (varsity athlete vs. nonathlete). Subsequently, George et 23 al. (1992) provided evidence towards clarifying this issue. They found that model ability was the most salient similarity cue among low-skilled female subjects, and that model sex was not a determinant of self-efficacy or performance. These findings suggest that the saliency of model characteristics may be dependent in part on the kinds of tasks being performed, as well as the ability level of the observer. W Verbal persuasion as a source of efficacy information usually involves receiving persuasory information from others. Bandura (1986) stresses that the effectiveness of persuasory information may be only as strong as the credibility or expertness of the person or persons providing the information. Judgments of personal efficacy are likely to change if the source of persuasive information is highly believable. Feltz (1984) argued that efficacy information obtained from verbal persuasion may also be derived from various cognitive techniques, such as positive imagery, self- talk, 'psyching-up', and performance deception. Though verbal persuasion is purported to be an influential source of efficacy information, little research in the sport literature has been conducted. One of the few studies to examine the effects of persuasion on performance beliefs was conducted by Ness and Patton (1979). Results showed that weightlifters were able to improve their maximum lifts when they were persuaded that they were lifting less weight than they actually pressed. Though self-efficacy was not assessed per se, beliefs about one’s capabilities to perform were altered, and performance improved. 24 .Strength performance has also been shown to be enhanced by other ”persuasive" cognitive strategies. Weinberg, Gould, and Jackson (1980) found that 'psyching-up” produced enhanced strength performance, but did not improve performance on tasks involving speed or balance. Similarly, Shelton and Mahoney (1978) found that competitive weightlifters facilitated performance by 'psyching-up." Again, these studies did not directly measure self-efficacy, but they provide some evidence that cognitive processes such as self— talk and imagery influence beliefs about motor performance, as well as actual performance. One major problem with the aforementioned research relates to the definition of 'psyching-up.’ The component elements involved in 'psyching—up' were not addressed. Thus, it makes it difficult to delineate the cognitive processes underlying the changes that occurred in performance. In an attempt to more directly assess the effect of cognitive strategies on strength performance, Wilkes and Summers (1984) had subjects employ one of several ”persuasive” strategies before attempting a strength task. Psyching-up and positive self-efficacy conditions produced the greatest changes in strength performance. In addition, subjects in these two conditions reported higher expectancy ratings for performance. Though the previous research has tended to show that certain “persuasive" cognitive strategies lead to enhanced performance, the underlying mechanisms mediating that change remain.unknownn In.other words, the previous research did not assess whether changes in self-efficacy accompanied changes in performance. In an effort to delineate this 25 relationship, Fitzsimmons, Landers, Thomas, and van der Mars (1991) replicated and extended the Ness and Patton (1979) study by assessing the effects of false information feedback on the self-efficacy and strength performance of experienced weightlifters. Subjects were provided with accurate. overinflated, or underinflated feedback regarding the amount of the weight they were attempting to lift. Results indicated that subjects who were told that they were lifting more weight than they actually were (false positive feedback) increased subsequent efficacy beliefs and weightlifting performance. However, self-efficacy beliefs were not found to be predictive of weightlifting performance. Instead, subjects' previous performance accounted for nearly all of the variance in subsequent performances. Thus, the study provided some support for Bandura's (1986) theory in that the hypothesized influential effects of verbal persuasion on self-efficacy were found to exist. However, changes in self—efficacy were not found to predict subsequent performance. Feltz and Riessinger (1990) conducted an investigation which examined the influence of in vivo emotive imagery and performance feedback on self-efficacy and muscular performance. Subjects were assigned to one of three conditions: mastery imagery plus feedback, feedback alone, or control condition. Results indicated that subjects in the imagery and feedback group reported significantly higher and stronger self-efficacy beliefs after each performance trial relative to the subjects in the other two conditions. Performance results also indicated that subjects in the imagery and feedback group outperformed subjects in the other 26 groups, but only on the first performance trial. Feltz and Riessinger (1990) also assessed the sources of subjeccs' efficacy beliefs, in an effort to determine whether the in vivo emotive imagery was operating as a influential source of efficacy information through verbal persuasion, as was previously suggested. Subjects reported that self- efficacy beliefs were primarily based upon their own performance accomplishments, as Bandura would suggest. Only a small percentage of subjects indicated that their efficacy beliefs were based on persuasory information. Thus, the strength and effectiveness of verbal persuasion as a source of efficacy information remains unclear. Emotional—Arousal Emotional arousal also serves as a source of efficacy information. According to Bandura (1986), arousal influences efficacy expectations through the cognitive appraisal of the arousal. If one interprets increased arousal as facilitory, personal efficacy and performance are likely to be enhanced. Likewise, efficacy and performance will suffer when arousal is labeled as fear or anxiety. A number of factors may influence one’s appraisal of arousal, including the source and level of the arousal, one's past experiences on the effects of arousal on performance, and the circumstances under which the arousal is elicited (Bandura, 1986). Bandura (1986) also posits that self-efficacy reciprocally influences subsequent arousal. In other words, emotional arousal is viewed as both a source of efficacy information and a co-effect of behavior (Feltz, 1984). Thus, one's assessment of emotional arousal partially determines 27 personal efficacy, which in turn influences future assessments of arousal. Research examining the arousal/self—efficacy relationship has provided mixed results. Feltz (1982) examined physiological arousal as source of efficacy information by measuring the heart rate of subjects prior to performing. Physiological arousal was not found to significantly influence diving efficacy. Similarly, self- efficacy was not found to be related to heart rate prior to subsequent dives. Thus, Bandura’s predictions were not supported. In an effort to further examine the arousal-self— efficacy relationship, Feltz and Mugno (1983) replicated and extended the Feltz (1982) study by assessing subjects’ perceived arousal as well as their actual physiological arousal (heart rate). Again, heart rate was not found to influence diving efficacy, nor did self-efficacy significantly affect subsequent heart rate. However, perceived autonomic arousal was found to be significantly related to self-efficacy on all four diving attempts. The authors found that lower levels of perceived arousal corresponded to higher levels of self-efficacy. Similar .results were found on the first of two dives for both male .and female performers in a subsequent study (Feltz, 1988b). 'fhese latter findings support Bandura's assertion that it is IMJC arousal per se that influences self-efficacy, but one’s .assessment of arousal that provides efficacy information (Bandura, 1986) . However, Bandura's hypothesis of a 11y:iprocal relationship among self-efficacy and physiological arrnisal has yet to receive support in the motor performance 28 literature. Bandura (1986) also notes that physiological sources of efficacy information are not limited to perceptions of emotional arousal. Cognitive appraisal of one's physiological state may also influence percepts of self— efficacy. For example, levels of fatigue, fitness, and pain may be perceived as signs of inefficacy (Feltz, 1988a). This may be especially relevant in tasks involving strength or endurance. Interpreting fatigue, pain, or lowered stamina as a sign of declining physical capacity is likely to lead to lower percepts of efficacy, and reduced effort and persistence in an activity (Bandura, 1986). Indeed, Taylor, Bandura, Ewart, Miller, and DeBusk (1985) found that heart attack patients tended to base their perceptions of cardiac capability on such observable signs as fatigue, shortness of breath, pain, and decreased stamina. These perceptions were found to be strong determinants of the patients' level of activity. Mood states have also been shown to influence self- efficacy beliefs. Research shows that positive mood states lead to higher judgments of capabilities than do neutral mood states, and negative mood is related to lowered efficacy expectations (Kavanagh & Bower, 1985). Bandura (1986) notes that the impact of mood on self—efficacy is not specific to the task at-hand. but pervades other achievement domains as well. Unfortunately, most of the motor performance research on physiological states has focused on the impact of arousal on self—efficacy. Virtually no attention has been given to other physiological states. Only one motor performance study E C. .9 ¢ .._ J . a 5 a A s « , ,3 5 S v . 3 . 2 . .. r a a . .3 S. .. 29 has (examined the effect of mood state on self-efficacy and performance. Using a handgrip strength task, Kavanagh and Hausfeld (1986) found no consistent effect for induced happy or sad mood states on subjects' efficacy expectations. Support was found for self-efficacy theory in that efficacy expectations were found to be a significantly correlated with strength performance. Moreover, mood was found to have a significant impact on handgrip performance. However, this relationship was not mediated by subjects' efficacy beliefs. The authors did find that mood significantly influenced subjects’ efficacy for performing a strength task in the future. Subjects in a positive mood state believed they could perform more push-ups than subjects in a negative mood state. However, these latter findings should be viewed with caution because subjects were never required to actually perform the task. Taken together, the findings on the impact of mood states on self-efficacy are equivocal at best. Given the paucity of research in the motor performance literature, the effects of fatigue, pain, physical condition, and/or mood on efficacy expectations and subsequent performance remains unclear. J . . E 3 JE- EE' Much of the literature examining the influential effects of various sources of efficacy information have provided support for Bandura's theory in that individual efficacy expectations for motor performance have been found to be partially dependent upon performance accomplishments, 3O vicarious experiences, and to a lesser extent verbal persuasion and physiological arousal. However, the majority of the research has failed to examine the causal and directional relationships posed by Bandura's model. Instead, the research has inferred causality and directionality from correlational relationships. Several studies have been conducted in an effort to directly examine the causal relationships in Bandura's model, which hypothesizes that the effects the four sources of efficacy information exert on performance are mediated through one's percepts of self- efficacy. Feltz and colleagues (Feltz, 1982, 1988b; Feltz & Mugno, 1983) have provided the most comprehensive attempt in the motor performance literature at delineating the causal mechanisms and directionality in self-efficacy theory. Feltz (1982) employed path analytic techniques to compare Bandura’s model of self-efficacy with an anxiety—based model in which self-efficacy was predicted to be an effect rather than a cause of performance. Subjects performed or attempted to perform four back-dives. Self—efficacy and physiological (heart rate) measures were taken just prior to each of the four attempts. Results were not very supportive of either model. Instead, self-efficacy was found to be the strongest predictor of back-diving performance only on the first trial. Self-efficacy was a significant predictor of subsequent diving performances, but previous performance was found to be a stronger predictor over the last three trials. However, self-efficacy and performance were found to be reciprocally related, as predicted. Heart rate was not found to predict self-efficacy, nor was self-efficacy found to be a predictor ‘|9 - 5 31 of heart rate. These findings led Feltz to respecify Bandura’s model to include previous performance accomplishments in addition to self-efficacy beliefs as direct predictors of performance. Feltz and Mugno (1983) replicated and extended Feltz's (1982) original study by adding autonomic perception as a measure of physiological arousal. Using the original protocol, the authors found that self-efficacy was the strongest predictor of back-diving performance, but once again only on the first trial. Subsequent trials were best predicted by previous back-diving performance, though self— efficacy was also a significant predictor on three of the four trials. Moreover, self-efficacy and performance were found to be reciprocally related. In addition, perception of autonomic arousal was found to be a better predictor of self-efficacy than was heart rate. However, the reciprocal relationship hypothesized to exist between self-efficacy and arousal was not found. Findings from the study provided further evidence that previous performances are important and direct determinants of future performance, and thus supported Feltz's respecified model of self-efficacy. Both-of the previous causal examinations of Bandura's model employed female subjects. Because existing literature suggests that males and females differ in their perceptions of self-confidence, fear, and arousal (Borkovec, 1976; Maccoby & Jacklin, 1974), a third study was conducted in order to assess possible gender differences in the causal mechanisms of self-efficacy. Feltz (1988b) employed the same back-diving task used in the previous studies, and tested her respecified model of self-efficacy using both male and female 32 subjects. Path analysis results revealed that the respecified model did not explain much of the variance for males or females, though it did fit the data better for female subjects. Specifically, for females self-efficacy was the strongest predictor of the first dive, but previous performance was the better predictor on a subsequent dive. Moreover, self-efficacy and performance were reciprocally related. Surprisingly, nothing predicted males’ first diving attempt. In contrast to consistent earlier findings, both self-efficacy and previous performance predicted the second diving performance, but the differential effects of the two variables were not as divergent. Self-efficacy was nearly as strong a predictor of performance as was previous performance. Also, contrary to prediction self-efficacy and performance were not found to be reciprocally related in the first back—diving performance. Overall, the results for females tended to support Feltz's respecified model. That is, after the initial performance of a novel task, previous performance accomplishments strongly influence subsequent performance attainments. However, this relationship was not as strong for male subjects. Feltz (1988b) indicated that the lack of (a significant self-efficacy-performance relationship on the first diving attempt may stem in part from a tendency for males to overestimate their efficacy beliefs. When performance 'scores were compared with efficacy scores, males' efficacy beliefs were nearly twice as strong as females' efficacy expectations. even though the two groups exhibited similar performances. This explanation is consistent with ‘10 a» —. .— 33 other motor performance research in which males have demonstrated an inflated sense of self—confidence (Feltz, Bandura, & Lirgg, 1989; Vealey, 1986). Several other studies have examined the causal relationships posed in Bandura's model of self-efficacy. In the motor performance domain, McAuley (1985) found that vicarious experiences (modeling) had a direct effect on self— efficacy and an indirect effect on performance through self- efficacy, as the model would predict. However, McAuley also found that vicarious experiences exerted a direct effect on performance. In fact, the modeling-performance path was stronger than the efficacy-performance path. Similar results were found by Schunk (1981) for math achievement among grade school children. However, Schunk found that the efficacy- performance relationship was stronger than the modeling— performance relationship. In an investigation of female basketball, soccer, and field hockey players, Haney (1991) found partial support for Bandura’s (1986) model. Path analysis revealed that objective past performance accomplishments were predictive of subsequent self-efficacy. Similarly, subjective ratings of past performance predicted self-efficacy, but only after the first performance trial. Anxiety was also found to be significantly related to efficacy beliefs. Self—efficacy was found to be significantly related to subsequent objective performance, but did not predict subjective ratings of performance. Contrary to Bandura’s theory, self-efficacy dified to, " I am concerned about my hitting in this >mpetition.' The CSAI—2 required subjects to indicate how iey felt just prior to hitting along a 4-point Likert scale. arceived arousal measures were employed because they have een shown to be better predictors of self-efficacy than have ore objective measures of arousal (Feltz & Mugno, 1983) . WW Objective erformance was measured by calculating each subject's ontact percentage. As reported by Courneya and Chelladurai 1991) , this type of hitting performance measure has been ‘ound to be the most responsive to skill execution and subtle :hanges in performance, and the least affected by :ontaminating factors such as random chance, an official's >ad call, or performance strategy. WW Subjects :ompleted a self-report hitting performance measure following each game (see Appendix C). The measure was designed to assess subjects' perceptions of their hitting performance. Subjects were asked to respond to the statement, “Please rate how well you think you hit in today's game". Responses were indicated on a Likert scale ranging from 1 (extremely poorly) to 10 (extremely well). W. Subjects were also asked to indicate how much effort they exerted while hitting during each game (see Appendix C). Subjects responded to the question, "How much effort did you put into hitting in today's game?" Responses were indicated on a scale divided into intervals of 10 percentage points, ranging from 0% effort to 100% effort. I, “-T' .7: 11‘": \ 39 Brncedura Permission was obtained from each of four head coaches prior to data collection. An explanation of the study was orally presented to players on each of the four teams, and informed consent was obtained from all players (see Appendix D). Responses to questionnaire items were kept confidential in that subjects were identified only by the last four digits of their social security numbers and the first letter of their last name. During a group meeting the week before the first data collection period, subjects completed the baseball self-efficacy questionnaire and the CSAI—2 in a "mock game" situation, so that all subjects were familiar with the questionnaires and understood how to complete the instruments prior to actual data collection. On each game day, subjects completed the hitting self- efficacy questionnaire just prior to the start of the game, and the CSAI—Z, the effort scale, and the subjective performance scale immediately following the conclusion of the game. In addition, the objective performance measure was calculated following each game. The questionnaires were administered by the same researcher for all games. For college players, questionnaires were placed in subjects' lockers prior to each game, and subjects completed the hitting efficacy questionnaire approximately 15-20 min before the start of the game. The questionnaires remained in each players' locker during the game, and were completed and placed in a sealed box at the conclusion of the game. Questionnaires were removed from the box after all players had vacated the 4O lockerroom. Questionnaires were also administered to the high school players approximately 15-20 min prior to each game. However, high school players completed the questionnaires while sitting in the dugout. After completing the hitting efficacy scale, the questionnaires were placed in a box until the completion of the game, at which time the remaining items were completed and the questionnaires were collected by the researcher. This protocol was followed for nine consecutive games for each team. W Prior to conducting the path analysis, correlation coefficients were computed for all independent variables in each wave in order to test for multicollinearity (r 2 .70) . In addition, correlation coefficients were computed for each .ndependent variable across the nine waves. These :oefficients were obtained in order to examine the stability f independent variables over time. Descriptive statistics were conducted for the aggregate E nine waves of responses for all subjects, as well as for 1e aggregate responses of the college and high school .ayer's. One-way analyses of variance were conducted on the gregate scores of each variable to determine whether the 0 groups differed on any of the variables in the model. fferences in the relationship among independent and Dendent variables for college and high school players were ;0 analyzed via z-t‘ests. Correlation coefficients were lputed, by wave and group, for each independent variable in 9 Pcm “IV 4- ._1 . :1 41 relation to the dependent variable it was predicting. Each coefficient was then transformed to a Fisher z statistic, so that the differences between the correlation coefficients of college and high school players could be tested (Glass & Hopkins, 1984). Group differences were tested in order to determine whether the two groups could be combined for subsequent analyses. The data were then analyzed using path analytic techniques. This analysis examined the fit of the hypothesized model to the data. Multiple regression equations were written for each of the hypothesized relationships presented in Chapter 1. Standardized path coefficients (Beta's) and zero-order correlations were calculated for each path in the model (see Figure 2). The hypothesized model was compared to a fully recursive model, which contained all possible paths to each dependent variable. The fully recursive model represented all the variance that could be explained by the independent variables in the model. The hypothesized and fully recursive models were compared using two techniques. First, a Chi-square goodness-of-fit statistic was computed (Pedhazur, 1982). Nonsignificant Chi-square values indicated that the data fit the proposed theoretical model, and thus provided a more parsimonious explanation of the data. The larger the jprobability associated with the Chi-square, the better the fit of the model to the data (Pedhazur, 1982). A second statistic was computed, the Q coefficient (Pedhazur, 1982), because the Chi-square test is affected by sample size. The Q coefficient represents the ratio of the ‘variance explained.by the hypothesized model relative to that 42 explained by the fully recursive model. Q can vary from 0 to 1, with values close to 1 indicating that the hypothesized model explains nearly all of the explainable variance in the dependent variables. x All... I knit»... bhuA [h (I! ’ 1 (I) CHAPTER IV Results This chapter is divided into two major sections. First, preliminary results relating to the overall responses of the subjects, as well as differences between the responses of the college and high school groups are presented. Group differences were examined in order to determine whether the responses of college and high school athletes could be combined for the path analytic procedures. Second, the results pertaining to each hypothesis are discussed in terms of the path analysis conducted on Bandura's model. The alpha level for all analyses was set at .05; however, actual p values are given when available so that significance levels can be reported more accurately. ]' . E 1 Descriptive statistics were conducted for the aggregate of nine waves of responses for all subjects, as well as for the aggregate responses of each of the two groups. One-way analyses of variance were conducted on the aggregate scores of each variable to determine whether the two groups differed on any of the variables in the model. Means and standard deviations for all variables for each wave are presented in Table 1 in Appendix E. Hi;;1ng_£e;fgzmance. As noted in Chapter 3, contact percentage was used as an objective measure of each subject's hitting performance. In addition to an objective measure of performance, subjective performance ratings (SPR) were also 43 I.“ E... o d an... M. A!9q bl - w.“ R»- V. .FA w.‘ ‘n I 44 employed as a measure of hitting performance. Overall, players' hitting performance was quite high (M = .825, SD = .140). On average, college players' hitting performance (M = .847, SD = .100) tended to be higher than high school players' performance (M = .805, SD = .168). However, a one-way analysis of variance indicated that this difference was not statistically significant, F (1,51)=1.22, p > .05. Figure 3 presents a comparison of the hitting performances of high school and college players. As can be seen, the pattern of hitting performances across the nine waves varied for the college and high school players. For example, high school players' hitting performances tended to fluctuate from game to game, whereas college players' hitting performances tended to improve across the nine waves (with the exception of Wave 7 hitting performance). An examination of the consistency of hitting performance across the nine waves for all subjects revealed that most of the correlations were moderate to low, indicating variability in performance. However, a number of significant correlations were found across the nine waves. For example. Wave 1 hitting performance was significantly related to Wave 2 (z,= .46, n,< .05). Wave 5 (r = .40, p < .03). and Wave 9 (I,= .45. n < .02) hitting performances. Similarly, Wave 3 hitting performance was significantly correlated with Wave 4 L; = .33, ng< .05), Wave 5 (r = .49, n,< .01), and Wave 9 (z = .36, n,< .05) performance. All intracorrelations for I hitting performance are presented in Table 2 in Appendix F. Self-report ratings of hitting performance tended to be rmmderate, with an overall mean rating of 5.35 (on a 10-point Contact Percent-go 1.0-l 45 —-9— College —o—- High School 0.7 I l I I I PFE G1 G2 CE G4 GS (36 G7 GB G9 Wovo Figure 3. Hitting Contact Percentage of College and High School Players by Wave. "(it In; fi-Tva‘i'. p- bk. V 6. Av C a u a .. Ce 4 '2 I H U . D F. L i « h N z- I v HA U u m. A.“ Eg‘ 46 scale). Both high school and college players reported moderate ratings of hitting performance (M = 5.51, SD = 1.43, M = 5.18, SD = 1.44, respectively). No overall differences were found for self-ratings of hitting performance. Subjective ratings of hitting performance were fairly consistent for the college players across the nine waves, but were quite erratic for the high school players (see Figure 4). A comparison of the actual and subjective performance scores reveals that both college and high school players' perceptions of performance and actual performance followed a similar pattern. Changes in subjective performance ratings tended to correspond with fluctuations in actual hitting performance. figlfi;fififica§y, Overall, the strength of self-efficacy beliefs for putting the ball in play was quite high (M = 84.68. SD = 13.61), ranging from 73.54 to 94.57 on a 100— point scale (see Appendix A). However, when comparing the two groups on self-efficacy aggregated across the season, college players reported significantly higher percepts of efficacy (M 91.41, SD = 10.12) than high school players (M = 78.66. SD 13.65). as indicated by a one-way analysis of variance, £(l,51)=14.63, p < .001 (see Table 16 in Appendix I for descriptive statistics of all self-efficacy items). As seen in Figure 5. college players' self-efficacy was consistently higher than high school players' efficacy across all nine waves. In addition, the self-efficacy of college players remained fairly consistent across all waves, whereas high school players' efficacy increased through Wave 4, and then fluctuated slightly for the remaining four waves. The self-efficacy intracorrelations across the nine “‘-.I\'-‘Ilc Subjootlvo Portormonoo Hotlng 47 7. —fl— College ‘ —o— High School 6 cl 5 - 4 -l 3 I I I I I I l T T I Hflfl§$l 61 G2 63 GM GB cm (N' 68 GB Wovo Figure 4. Subjective Performance Ratings of College and High School Players by Wave. Sou-Efficacy 48 1oo - —o— High School SE ——-o—- College SE 90 III 80 - 70 I I I I I I I I I O 1 2 3 4 5 6 7 8 9 Figure 5. Self-efficacy of College and High School Players by Wave. ‘1) t\) P: - 'l’) 1!) ‘VV o.. . \N a W 49 waves for all subjects are presented in Table 3 in Appendix F. As can be seen, self-efficacy was highly correlated across the nine Waves, and all correlations were significant at the 9,5 .001 level. The consistently strong correlations among self-efficacy, in contrast to the weaker hitting performance correlations, suggested that players' efficacy beliefs were quite resilient to fluctuations in performance. Anxiety, Anxiety was examined in terms of its cognitive and somatic components (see Appendix B). Overall, perceived cognitive anxiety (M = 16.61, SD = 4.30) was higher than perceived somatic anxiety (M = 13.31, SD = 3.01). A paired t test revealed that this difference was statistically significant, t(52)=7.70, p < .001. However, one-way ANOVAs conducted on the cognitive and somatic anxiety scores of the two groups aggregated across the season revealed that the somatic anxiety reported by college players (M = 12.23, SD = 1.60) was significantly less than that of high school players (n;= 14.27, 52 = 3.62).‘E(1,51)=6.70, p < .013. The two groups did not differ on aggregated cognitive anxiety. Cognitive anxiety tended to decrease over the season for both college and high school players (see Figure 6). Compared to Wave 1 levels, both groups reported lower levels of cognitive anxiety for all subsequent waves. However, as illustrated in Figure 6 cognitive anxiety fluctuated to some extent over the season and from game to game. After the initial wave, college players maintained a fairly consistent level of cognitive anxiety, with the exception of Wave 6. High school players reported slightly less consistent levels (of cognitive anxiety across the nine waves, but the changes in anxiety levels were small, fluctuating no more than four Cognitive Anxiety Level 19' 18-l 17- 15" 14 50 “—fl— College —+—- High School 61 (i G3 G4 G5 GS G7 GB G9 Gomo Figure 6. Cognitive Anxiety Levels of College and High School Players by Wave. GA} v'v‘ . O ”a f a... «In p a“... «an - l 4.5.. aby- duo. nu-nv RS. AA 91... aux. . ,I-q vvg. ‘A-o. nu... _ 'I~‘ v '43.: ~. .. ‘ Vcl... , '1-. l.~* I'D l . a ‘ "~ ~ ‘ .. "'1 ”I- VV “ ‘. v.14; 51 points (on a 36-point scale) over the nine waves. The intracorrelations for cognitive anxiety over the nine waves were quite strong, and all but one correlation was significant (see Table 4 in Appendix F). Wave 3 cognitive anxiety was not significantly related to Wave 9 cognitive anxiety (I = .30. p > .05). Similar to self—efficacy, the strong correlations among cognitive anxiety across all nine waves suggested that perceived cognitive anxiety was fairly stable from game to game. Figure 7 presents the pattern of somatic anxiety for college and high school players across the nine waves. Though lower levels of somatic anxiety were reported by college players, both groups indicated game to game fluctuations. As can be seen, college players’ level of somatic anxiety tended to decrease across the first six waves. An increase in Wave 7 somatic anxiety to a level similar to that found in Wave 1 was again followed by slight decreases in Waves 8 and 9. High school players' somatic anxiety tended to rise and fall on a game-to-game basis. Thus, an undulating pattern of somatic anxiety was reported by high school players, whereas college players tended to report more consistent levels of somatic anxiety. The intracorrelations for somatic anxiety were also very different than those for cognitive anxiety. As shown in Table 5 in Appendix F, no consistent relationship was found for somatic anxiety. Somatic anxiety in Waves 1 and 2 were highly correlated, but did not correlate with somatic anxiety in Waves 5 through 9. However, a fairly strong relationship existed for somatic anxiety in Waves 3-9. as 16 of the 21 Somotlc Anxiety Level 16- 15- 14. 13-1 12' —0— College —0— High School 11 - — u-l u: d .1 d d I G1 G2 63 G4 GS GB G7 GB 69 Wave Figure 7. Somatic Anxiety of College and High School Players by Wave. 53 intracorrelations were significant. Effigzt. On average, subjects reported exerting a relatively great amount of effort while hitting. On a 10- point scale, effort ratings aggregated across the season ranged from 4.75 to 10.00 (M = 8.51, SD = 1.36). A one-way ANOVA indicated a significant group difference in aggregate ratings of effort, E(1.51)=9.96, p < .003. College players reported exerting more effort while hitting (M = 9.09, SD = 0.71) than high school players (M = 7.99, SD = 1.60). An examination of effort across the nine waves suggests that college players consistently thought they exerted higher levels of effort while hitting compared to high school players. As seen in Figure 8, college players consistently reported an effort rating of 8.5 to 9.5 for all nine waves. Conversely, high school players’ ratings fluctuated from game to game, and did not exhibit a consistent pattern. Following the lowest level of effort in Wave 2, high school players' effort consistently increased across Waves 3, 4, 5, and 6, at which point lower levels of effort were reported for the final three waves. Intracorrelations for effort were somewhat consistent, ranging from .02 to .78 (see Table 6 in Appendix F). Wave 6 effort differed from the effort exerted in most of the other Waves. and was found to be significantly related to only Wave 2 and Wave 7 effort. Most of the other intracorrelations for effort were significant, suggesting that the amount of effort exerted while hitting was similar across the nine waves. Effott 54 10" --I-- Ckmege —9— High School 9 .1 8 I: 7 I I I I I I I I I (31 (32 <33 <34 (35 (36 (37 (38 GB Wovo Figure 8. Perceived Effort of College and High School Players by Wave. 55 J . i 3 . 1] Pearson product-moment correlations were conducted in order to determine whether multicollinearity (r 2 .70) existed among any of the independent variables. The correlations for independent variables existing in different waves (e.g. Wave 1 cognitive anxiety and Wave 2 somatic anxiety) were not examined because only first-order dependence was considered in the path analysis. That is. direct paths were not tested across waves. Therefore, highly correlated variables in different waves were not perceived to be problematic. Table 7 in Appendix G presents the intercorrelations for the independent variables for each Wave. As can be seen, cognitive anxiety and somatic anxiety were highly correlated in Wave 3 (1,: .70, p < .001), Wave 4 (L = .68, p < .001) Wave 6 (r = .70, p < .001). Wave 7 (r = .67, p < .001). and Wave 8 (1,: .72, p,< .001). In addition, the correlation coefficient for objective hitting performance and self- efficacy was very high in Wave 9 (r = .70, p < .001). These correlation coefficients suggested that multicollinearity may have existed in these waves, and thus problems may have arisen in the estimation of regression statistics. In an effort to circumvent potential problems with highly correlated independent variables, multicollinearity was also tested by examining the tolerance estimates of the predictor variables in each regression equation. Tolerance estimates are a measure of the degree of linear association between the independent variables. Small values for the tolerance of a variable indicate that the variable is, or almost is, a linear combination of the other independent 56 variables. Variables with small tolerances (less than .01) were not allowed in the regression analyses. However, all tolerance values were found to be acceptable and no independent variables were removed from the regression analysis. W Group differences pertaining to the relationships between independent and dependent variables across the nine waves were examined via z-tests. Zero-order correlation coefficients were computed by wave and group for each path associated with the hypothesized model. Each coefficient was then transformed to a Fisher z statistic. In doing so, it was possible to test the differences between the correlation coefficients of college and high school players for each hypothesized path (Glass & Hopkins, 1984). One-hundred and seventeen pairs of correlation coefficients were examined, and only four significant differences were found. All correlation coefficients for the hypothesized paths are presented in Table 8. The correlation coefficients associated with the self-efficacy --> subjective performance paths were found to be significantly different in Waves 5 (z_= 2.29, p < .05) and 9 (z = 2.23, p < .05). In Wave 5, college players reported a significant positive relationship (I = .47, p_< .03), whereas high school players exhibited a negative relationship (1 = -.26, p > .10). The opposite association was found in Wave 9. High school players demonstrated a non-significant but positive relationship between self—efficacy and subjective performance (I = .21, pl> .10). On the other hand, self-efficacy was 57 oocoEHOuqu assays: u uwm huwfixco OflumEOm u xcceom >ocowuuoiuaom H mm xuoflxco o>wuficmoo u xcomou ouccEHOMqu o>fiuuofln3m u mam oocoEMOHHoQ mewuuwn mzofl>oHQ H mm Ho. V Q «« mo.vo.. .om>u Cacao n muoxcaa Hoozom saws hoax» UHOQ u muo>oHQ omoHHOO "ouoz «amm.i «vv.i «vv.: «0.: Hm. mm.i «bv.i ma. hm.i mo. wm.u eawo.i wn.n am.i ho. h«.i no. cuvh.i mm.?i xfiéhfi «tvm.i um.) «gwm.i hm.i mo. hm.n «Hm.i ov.i ov.i HH.: mn.n «v.1 on.i ov.i oavm.u chv.i oH.i Gm.i mm Aliosadou «omw.u ma. oavm.i mo. va.i mo.i vm.i mv.n ov.i wo.u mo. mo. Ho. mH.i umv. an.u ov. om.i inEom.?i mom mm.i av. na.i mm. mo.- ma.i om.i Ha. oom.i mu. wv.u mm. mm.i cvv.i Hm. ca. ma. hm.u xcoEom.?i.mm ivm.u mo.i ioam.u ma. mo. mo.i om.i «mm.i «bv.i ma. mo. on.i «a. va. vm. Ho.i amm. ca.u XGc3 .Qsouo 0cm w>m3 >9 HOUOZ pmwflmmzanxm 05D MOM mucmfloflwuwou COHDMHOHHOU HODHOIOMON .m onooe HO. V Q it mo. V Q t .oaku sedan u muoxcaa Hoozom sow: “waxy UHOQ u mumhoam ommaaou "muoz 58 mv. Om. «mm. tOm. mm. Hm. kcbm. HO. «om. MO. mH. OO. NO. Gfim. Om. OM. IO‘. MM. mmm.0:.UHOuEm HN. vH. VO. OH. mN.! HN. OH.! OO.! ON. Cflm.l mO.I OM. MN. chi. "0.! m0. mH. OO. mmm.o:.mm «mv. mm. vH.! 60.! OH. HO.! OV. HO. mfl. MH.I MH.! GN.! Hm. GINO. th‘. "0.! Hfi. OH. uflz.¢:.UHOwEm mN. vv. HO. 0v. «Nm. smv. mv. ##wm. tsmm. ON. NH. NO. $0.! hm. mH.! OH. OH.! OO‘. me.?l mm ma. cwv. mo. Nm. ON. ON. 6v. NH. vv. ON.! NM... Cam. HH... CHm. ON.! «0.! NH. FN. UHOuwm all rfl tNm. vH. mv. ON. ON. Nm. OO.I Om. mN. HH.I $0.! mO.I Cflm. mO.! 0°. OH. OH. hm. mm 6:.mmm cimb. VO. *«vw. mv. ¢®m. «mm. scam. scab. NN. HO.! 0“. mm.! Om.! OO. 50. OH. ON. HN. mm 8:.mm Suflm O O P m m v m N H w>o3 .Ap.ucouv m mange 59 negatively related to subjective performance for college players (2 = -.54, p < .02). The correlation coefficients associated with somatic anxiety and self-efficacy were also found to significantly differ in Wave 9. High school players exhibited a moderate. negative relationship (r = -.58, p < .02). whereas virtually no association between somatic anxiety and self-efficacy was found for college players (r = .08, p > .10). Lastly, the correlation coefficients associated with the previous hitting performance --> somatic anxiety path were found to significantly differ in Wave 8. Previous performance was negatively related to somatic anxiety for college players (L = -.46, p,> .05), and positively related for high school players (r = .41, p > .05). No other differences were found. It should be noted that the z-ratios reported were compared a critical z-value of 1.96. However, the probability of making a type-I error was greatly increased due to the large number of comparisons being made. Thus, the Bonferroni t,(Miller, 1966) method was employed to estimate the maximum probability of making a type-I error. When the critical z:values were adjusted to account for the multiple comparisons, none of the correlation coefficients were found to significantly differ. Thus. because no consistent pattern of group differences emerged from the z-tests. the two groups ‘were combined for all subsequent analyses. W Five hypotheses pertaining to the causal relationships in Bandura's model were set forth (see Chapter 1). In order tn) test these hypotheses, path coefficients were calculated 60 via multiple regression analyses for each of the hypothesized relationships. Tables 9 to 13 contain the path coefficients for the hypothesized model. Each table contains only the path coefficients for equations that tested a particular hypothesis. A complete representation of the hypothesized and full model is provided in Tables 14 and 15 in Appendix H. Based on the lack of significant differences found in the relationship between independent and dependent variables of college and high school subjects, the scores for both groups were combined and used in the regression analyses. The first hypothesis posited that past hitting performance would predict somatic and cognitive anxiety in the theoretical model. Hitting performance was predicted to be negatively related to anxiety. Multiple regression analyses revealed virtually no support for the past performance--> anxiety relationship. As seen in Table 9. actual hitting performance did not predict somatic or cognitive anxiety in any of the nine waves. On the other hand, the subjective performance rating ——> somatic anxiety path coefficient was significant in Wave 9 (E = -.387, p < .05), and was marginally significant in Wave 3 (E = -.311, p < .07) and Wave 7 (B = -.393, p < .07). Likewise, subjective performance significantly predicted cognitive anxiety in Wave 7 (B = -.514. p < .01). and was a marginally significant predictor of cognitive anxiety in Wave 3 (fl = +.296, p < .09). In all five cases, significant path coefficients were in the predicted direction. Better performances were related to lower levels of anxiety. and poorer performances were associated with increased anxiety levels. The second hypothesis was related to the variables 61 mocofiu0muom msofl>oum uo mmcwuou o>wuoonnsm u mam huoflxco ofiuoEOm n XGMEom >umflxco o>fiuflcmoo u Kenmou oucoEHOMHom ocwuufin msofl>oua H mm HO. VG it. mo. v a . OH. v Q + .amm.- oao. +mmm.- meo. mao.- mmm. +Ham.- mam.- NOH.- x:<&om Ail mom Heo.- GHH. «mo. moa.- mmm.- «mo. mmo.- mea. HAH.- xcoeom A-- no HHm.- mmo. ..efim.- mmH. mmm. ova. +mmm.- mmo.- mmH.- xcamoo A-- mom mam.- mmH.- amo.- Hao.- omm.- mmo.- mmo.- moo.- moo. xcomoo A.. no homo m m a m m o m m H m>o3 .H mflmonuoohm HOM mucoAOMMwooo comm .m mange 62 influencing self-efficacy beliefs. Hypothesis 2 stated that subject’s past performance experiences and anxiety would significantly predict self—efficacy. Moreover, past hitting performance was hypothesized to be positively related to self-efficacy, whereas anxiety was expected to be negatively related to self—efficacy. Path coefficients for Hypothesis 2 are presented in Table 10. The results of multiple regression analyses revealed strong support for both portions of the second hypothesis. The cognitive anxiety —-> self- efficacy path coefficients were significant in Waves 2 (E = - .433), 4 (E = -.485). and 6 (E = -.391). and were marginally significant in Wave 3 (E = -.399). Moreover, somatic anxiety predicted self-efficacy in Waves 1 (E = -.434), 7 (E = - .471), and 8 (E = -.599). As predicted, both cognitive and somatic anxiety were found to be negatively related to self- efficacy. Strong support was also found for the influence of objective past performance on self-efficacy. The hitting performance --> self-efficacy path coefficients were significant in Wave 2 (E = .434), Wave 3 (E,= .343). Wave 5 (E = .470), Wave 6 (E = .441). Wave 7 (E = .493) and Wave 9 (E = .632). All significant paths were in the predicted direction, as stronger hitting performances were associated with higher percepts of hitting efficacy. Conversely, subjective ratings of prior performance predicted self—efficacy in only two waves. A significant SPR --> self-efficacy path coefficient was found in Wave 1 (E = .308. p < .05), and the SPR --> self-efficacy path coefficient was marginally significant in Wave 2 (E = .255, p < .10). Thus, despite the strong predictive ability Of 63 mcwumu oucofiu0muoa o>fiuumnnsn u mam mucoEHOwqu mason“: msofi>ona n ma mo. V m 4. OH. V m + OHO. mmo. OMN.! mmH. mmO.! PMH. NMO.! +mmN. «mom. mm Al! mam «xNMw. VPH. «kmmv. «savv. «Ohfi. OHN. smvm. sawmv. vvo.l mm Al! mm NmH.I «mam.l «Hbv.! moo. omO. HOO. mmH.! NHN. #de.! mm Al! XSdEOm ONH.! mNH. bOH.! #Hmm.! OOH.I immv.! +mmm.l smmv.l bwo.l mm Al! XC4OOU Suwm O O F m m v m N H m>63 .m mflmonu0Q>m no“ mucoAOANmooo zoom .OH manna 64 actual hitting performance, the data provided little support for the subjective performance --> self-efficacy relationship. In the third hypothesis. be the only predictor of effort in the proposed model. self-efficacy was purported to Results of the path analysis indicated that self-efficacy was at least a marginally significant predictor of effort in six of the nine waves, thus providing support for Hypothesis 3. As seen in Table 11, the self-efficacy --> effort path coefficients were significant in Waves 1 (E = .436), 3 (E = .393). and 8 (E = .414), and marginally - .278). All (E - .313) and 7 (E - .372). 5 (13 = significant in Waves 2 paths were in a positive direction, indicating that stronger percepts of efficacy were predictive of higher effort expenditure while hitting. That self-efficacy significantly predicted effort in only four of the nine waves suggested that variables other than self-efficacy were influencing the amount of effort exerted while hitting. This possibility was tested by examining the fully recursive model over all nine waves (WhiCh regressed effort on previous performance. anxiety. and self—efficacy) and comparing it to the hypothesized model via a Chi—square goodness-of-fit test and a 'Q' ratio (Pedhazur, l982 ) The x2 goodness-of—fit test resulted in a SL1ggeSt:ing that the hypothesized model adequately explained th . . e relationships among the variables. However, as a further comparison of the full and hy pothesized models, a Q-coefficient was computed in order to 65 sooonuuo-uaom u mm Ho. v a .. mo. v a . oa.vQ + amm. .eaq. .mam. cam. .mmm. won. .mam. +mam. ..mme. onouum A-- mm sumo a m A o m o m m a o>63 .m memonuoahm qu mucofl0fluwooo sumo .HH canoe 66 examine the differences in the variance explained by the hypothesized model relative to the variance explained by the full model. As noted in Chapter 3, the Q coefficient may vary from 0 to 1, with values close to 1 indicating that the hypothesized model fit the data and can explain nearly all of the ”explainable" variance in the dependent measures. The Q coefficient for the hypothesized model was only .21, suggesting that much of the variance was left unexplained. An examination of the full model paths revealed that with the exception of Waves 2, 6, and 7, self-efficacy was the only predictor of effort. Somatic anxiety (E = -.540, p < .05) and cognitive anxiety (E = .525, p < .05) significantly predicted effort in Wave 7. In addition, the subjective performance --> effort path coefficient was marginally significant in Waves 2 (E = -.341, p < .10) and 6 (E = -.353. p < .06). Surprisingly, no other variables in the full model were found to predict effort. Apparently, the addition of four paths was enough to account for a significant amount of variance in effort even though no individual variable exerted a strong influence on effort. Given that other variables predicted effort in only two of the nine waves, it appeared as though self-efficacy was the strongest, if not the only predictor of effort in the model. Hypothesis 4 stated that self-efficacy and effort would be the only predictors of performance, and that self-efficacy would be the stronger of the two predictors. Moderate support was found for this hypothesis in that self-efficacy was a significant predictor of actual hitting performance in Wave 1 (E = .449, p,< .02), Wave 5 (E = .379, p < .02), and .349, p < .05). In addition, the self-efficacy - Wave 8 (E 67 -> performance path coefficients were marginally significant in Wave 2 (E = .339, p < .06) and Wave 3 (E = .310, p < .08). In contrast, the effort -—> hitting performance path coefficient was found to be significant in Wave 5 only, and this effect was marginal (E = .296, p < .07). All hypothesized model path coefficients for Hypothesis 4 are presented in Table 12. That self-efficacy was at least a marginal predictor of actual performance in five of the nine Waves supports the hypothesis that self-efficacy is the stronger predictor of performance. However, in order to test the hypothesis that self-efficacy and effort are the only predictors of performance, the hypothesized model was compared to the fully recursive model. The X2 goodness-of-fit test was significant. E2 (36, n=38) = 69.36, p < .05, and the Q coefficient was .13, indicating that variables other than self-efficacy and effort were accounting for the variance in hitting performance. An examination of the full model indicated that there was a trend toward significance for the previous performance -—> performance path coefficients in Wave 2 (E = .420, p,< .08) and Wave 4 (E = .311, p < .08). In addition, somatic anxiety (E = —.432, p < .03) was a significant predictor of performance and cognitive anxiety (E = .325, n,< .06) was a marginally significant predictor of performance in the first wave. Somatic anxiety was also a marginally significant predictor of performance in Wave 2 (E = -.412, p < .06). However, only somatic anxiety exhibited an inverse relationship with performance. Thus, in addition to self-efficacy, previous performance and anxiety also 68 onwuou mucoEHOHHoQ o>wuoofl95m u mam oucoEMOMMOQ mcwuuwn Hanuoo u yam huoofiuumuuaom H mm Ho. vfiu.. mO.VQ.¢ oH.VQ+ +mvm. +mmm. +mam. mvm. .moo. +vom. .«bmm. mom. .mmm. mmm All uncuum Hom.! moo.l oHo.- moo. Hmo. oHo.i +Nom.! bmo. moo. mom All mm omm. oam. SSH.- woo. +omm. moa. woo. moo.! moa. box A)! uncuum MHH. «mom. Hmo. +omm. «mom. mam. +on. +mmm. .mov. Dem Ali mm homo m m r m m v m m H w>m3 .v mflmonuoa>m MOM mucoHOHmuooo some .NH manoe 69 exerted direct influences on actual performance. The relationship between self-efficacy. effort and subjective performance was quite different than the The self- relationship exhibited with actual performance. efficacy --> subjective performance path coefficient was - -.262, p < .10), marginally significant in Wave 3 only (E and contrary to prediction. the relationship was inverse. on the other hand, significantly predicted subjective Effort. performance in Wave 1 (E and Wave 5 (E = .445) . There was also a trend toward significant effort --> subjective performance path coefficients in Wave 4 (E = .312, p < .08), Wave 8 (E = (13 = .342. D < .08). .398), Wave 3 (E = .557), (E = .304, p < .07), Wave 7 .08). and Wave 9 All of the .332 , D < paths were in the predicted direction. The amount of error associated with the prediction of subjective performance in each of the nine waves was quite large, again suggesting that variables other than self- efficacy and effort were not accounted for in the prediction Thus, the fully recursive model. of subjective performance. Which included past objective and subjective hitting performance. somatic and cognitive anxiety, self-efficacy. and effort as predictors of performance, was again compared The X2 value for subjective to the hypothesized model. _ = 38) = performance was found to be significant. 2&2 (36. n indicating that the hypothesized model did 89°47. g< .01, The Q coefficient was also extremely low n . 0: fit the data. ( - . - O7 ) providing further support that the hypotheSized model di d not explain much of the variance in the dependent ' V . ¥_ 70 An examination of the full model indicated that in addition to effort and self-efficacy, previous subjective performance significantly predicted subsequent subjective performance in Wave 4 (E = .399, p < .05). Wave 6 (E_= .460, p < .01) and Wave 9 (E = .462, p < .05). In addition, the cognitive --> SPR path coefficient was significant in Wave 6 (E,= -.773, p < .01). and the somatic anxiety --> SPR path was marginally significant in wave 2 (E = -.360, p < .10). As predicted, both cognitive somatic anxiety were inversely related to performance. Though previous subjective performance and anxiety accounted for a portion of the variance in subjective performance, effort still significantly predicted subjective performance in five of the nine waves in the full model. Thus, the influence of effort on subjective performance appears to be strong even though it does not account for all the variance. Hypothesis 5 stated that subjective performance would be a stronger predictor of self—efficacy than would objective performance. No support was found for this hypothesis in that the subjective performance —-> self-efficacy path coefficient was significant in Wave 1 (E = .308, p,< .05) only. As noted in Hypothesis 2, objective performance was a significant predictor of self-efficacy in six of the nine waves. Thus, contrary to prediction, the hypothesis that perceptions of performance, as opposed to actual performance. more strongly predict subsequent efficacy expectations was not supported by the data. Table 13 contains the path coefficients for Hypothesis 5. mcfiuou oucoEHOuuoa o>wuumwnsm u mam hooufiuumiuamm H mm moanuOuqu onwuufln Hozuoo n was 71 Ho. v o .. mo.Vme OH. v o + 0H0. mmo. mm~.- man. mmo.- oma. mmo.- +mmm. .mom. mm A-- mom ..mmo. «SH. ..mmo. ..Hoe. .oao. cam. .mom. ..ome. 440.- mm A-- on: S a a...“ o>o3 .m mflmonuoamm MOM mucofloflumooo numm .mH magma 72 W Hypothesis 1 stated that past performance accomplishments would predict subsequent somatic and cognitive anxiety. Virtually no support was found for this hypothesis in that objective performance did not predict anxiety in any of the nine waves and subjective performance predicted anxiety in only two waves. Hypothesis 2 stated that past performance and anxiety would be significant predictors of self-efficacy. Moreover, past performance was hypothesized to be positively related to self-efficacy, whereas anxiety was predicted to be negatively related to self-efficacy. Strong support for Hypothesis 2 was found in that cognitive and/or somatic anxiety predicted self-efficacy in seven of the nine waves, and hitting performance predicted self-efficacy in six of the nine waves. Subjective performance was not found to be a predictor of self-efficacy. The third hypothesis received moderate support from the hypothesized model. The selfeefficacy -—> effort path coefficients were significant in four waves and marginally significant in two other waves. An examination of the full model revealed that no other variables were consistent predictors of effort. Moderate support was found for the fourth hypothesis. Self-efficacy predicted hitting performance in five of the nine waves, but effort predicted performance in only one wave. Conversely, effort predicted subjective performance in three waves and the path coefficients were marginally significant in four other waves. No significant self- 73 efficacy —-> subjective performance path coefficients were found. CHAPTER V Discussion The primary purpose of this study was to examine the causal relationships in Bandura’s (1977, 1986) theory of self-efficacy in a naturalistic sport setting over an extended period of time. Based on a review of the sport and motor performance literature relating to self-efficacy theory, baseball players' percepts of hitting efficacy were hypothesized to be the strongest predictor of effort expenditure while hitting, as well as actual hitting performance. Likewise, hitting performance was hypothesized to be the strongest predictor of self-efficacy. Furthermore, players' perceptions of hitting performance were hypothesized to be stronger predictors than actual hitting performances of self—efficacy beliefs. Finally, cognitive and somatic anxiety were hypothesized to be predictors of self-efficacy. 1' . E 1 Before discussing the results of the path analysis, some of the preliminary results warrant further discussion. In particular, the findings relating to the high intracorrelations among self-efficacy over time seem particularly salient. The self-efficacy of all players was found to be consistent over the nine waves, despite fluctuations in performance. This finding is important for two reasons. First, it provides some support for Bandura's (1986) assertion that percepts of self—efficacy are not merely reflections of past performances, but beliefs in one's 74 75 capabilities to execute specific actions. Had players' efficacy beliefs simply been reflections of their hitting performance, percepts of efficacy would have undulated in accord with the fluctuations in objective and subjective hitting performance. As shown in Figures in 3, 4, and 5, this was not the case. Self-efficacy remained fairly consistent, whereas performance fluctuated across waves. Second, the consistency of self-efficacy is noteworthy because it speaks to the resiliency of self-efficacy beliefs. Bandura (1986) posits that individuals who are fully assured of their capabilities maintain high percepts of efficacy even after encountering repeated failures. In the present study. all players tended to report very strong efficacy beliefs, though collegiate players consistently reported higher efficacy beliefs than high school players. In fact, players typically used only the upper quarter of the efficacy scale. Players maintained these strong efficacy beliefs despite fluctuations in hitting performances. These findings are consistent with those of Brown and Inouye (1978). who found that highly efficacious individuals remained efficacious even after repeated failure on an unsolvable problem. In conclusion, self-efficacy was found to be consistent across the nine waves, as well as resilient to acute changes in performance. The reasons for the resiliency of efficacy beliefs demonstrated in the present study remains unclear. Future research needs to investigate the underlying processes involved in the development of resilient efficacy expectations. For example, why are some individuals able to maintain their efficacy beliefs in the face or adversity. while others succumb to failure experiences? Can individuals 76 learn or be manipulated to remain efficacious after experiencing repeated failures? These and other questions should be addressed in future research. W The primary purpose of the present study was to examine the causal relationships in Bandura's (1977) model of self- efficacy. The overall results of the path analysis offered little support for the hypothesized model. Contrary to predictions in Hypothesis 1, past performance accomplishments did not predict cognitive or somatic anxiety, nor were they found to be the most salient source of efficacy information, as Bandura (1977, 1986) would assert. Similarly, subjective performance predicted anxiety in only two waves. Although several explanations may account for these findings, the most likely explanation pertains to the measurement of anxiety. Players' perceptions of anxiety were assessed retrospect- ively. That is. players completed anxiety measures following their hitting performance. Thus, players may have "adjusted" their assessments of anxiety to reflect their most recent hitting performance, rather than how they felt just prior to performance. Thus, it would not be surprising to find that hitting performance from a previous game was a poor predictor of anxiety. The effectiveness of assessing anxiety retrospectively should be a concern for future research endeavors. Past performance accomplishments and anxiety were also found to be consistently strong predictors of self-efficacy. as proposed in Hypothesis 2. The past performance --> self- efficacy path was found to be significant in Waves 2. 3. 5. 77 6, 7, and 9. Conversely, subjective ratings of performance predicted self-efficacy in only one of the nine waves. The past performance --> self—efficacy paths were not significant in all nine waves, and the lack of significant paths in several waves may have been due to methodological concerns. Prior performances consisted of each player's average contact percentage for the most recent game, and self-efficacy measures were completed just prior to each game. Typically, games were played every two or three days. This protocol allowed a considerable amount of time to elapse between each game. The time span between games may have allowed other factors to influence players' self-efficacy beliefs. For example, the potential effects of a poor hitting performance may have been offset by hitting instruction or batting practice. Perhaps players were able to maintain their percepts of hitting efficacy based on positive feedback from hitting practice. Players may also have engaged in cognitive strategies designed to circumvent the debilitating effects of poor performances. Research has shown that baseball players typically engage in a variety of cognitive strategies designed to facilitate their performance (George, 1988; Gmelch, 1972). It is reasonable to assume that players in the present study engaged in self-talk, :hnagery, relaxation, and other cognitive techniques in an effort to overcome or rebound from the effects of poor performances. Contextual factors may also have contributed to changes in self-efficacy, such as the opposing pitcher or team, or whether a player's team won or lost a previous game. This possibility is illustrated by examining the the contextual 78 factors associated with various games across the nine waves. For example, after reporting very strong efficacy beliefs in Wave 4 and winning their respective games, high school players' self-efficacy expectations dropped for Waves 5, 6, and 7, when they faced their toughest competition. This decrease occurred despite strong hitting performances in Waves 4 and 6 (see Figures 1 and 3). Conversely, after winning in Wave 7 despite extremely poor hitting performances, high school players' self-efficacy beliefs rose to their highest level in Wave 8. This increase in self— efficacy may be attributed to the weak opponent (pitching) each team faced in Wave 8. Thus, opponent's ability may have been a more salient source of efficacy information than previous hitting performance. The potential influence of contextual factors on college players’ efficacy beliefs was less clear, mainly due to little variation in self-efficacy across the nine waves. However, contextual factors may still have influenced efficacy expectations. For example,_players' self—efficacy dropped slightly in Wave 4 despite weak opposition in Wave 4 and strong hitting performances in Wave 3. Perhaps this decrease is partially due to each team losing their respective game in Wave 3. Similarly, efficacy expectations increased slightly from Wave 7 to Wave 8, even though players exhibited their lowest level of hitting performance in Wave 7. Again, winning the previous game may have accounted for the increase in efficacy beliefs. However, each team played a conference rival in Wave 8. Perhaps the rivalry led to additional preparation for the game, and thus accounted for the heightened efficacy beliefs. These factors as well as others ghoul: Cergamly the factors may 5 ccntribute CS The sec: wou‘. be a 5: relationship . .' ‘1 . port.a-iy Sb? arxiet was a somatic anxie All path coef findings lend that emotiona information. research indi< associated wi‘ 1988b; Feltz . The resu anxiety and s assessing cog bl I tree: . ‘Lcsqrs 0C arOl‘Sa‘ 0: ' e S Cogniti C;- sk‘lh "N . 03mins? 79 others should be investigated in future research endeavors. Certainly the present findings suggest that contextual factors may serve as sources of efficacy information that contribute to individuals’ appraisals of efficacy beliefs. The second part of Hypothesis 2 stated that anxiety would be a significant predictor of self-efficacy, and the relationship would be inverse. Results of the path analysis partially supported this part of the hypothesis. Cognitive anxiety was a predictor of self-efficacy in three waves, and somatic anxiety predicted self-efficacy in three other waves. All path coefficients were in the predicted direction. These findings lend support to Bandura's (1977, 1986) contention that emotional arousal can be a strong source of efficacy information, and also supports the findings from previous research indicating that lower levels of arousal are associated with higher percepts of efficacy (Feltz, 1982, 1988b; Feltz & Mugno. 1983; McAuley 1985). The results pertaining to the relationship between anxiety and self—efficacy also emphasize the importance of assessing cognitive arousal in addition to physiological indices of arousal. Players' perceptions of cognitive anxiety predicted self-efficacy just as strongly and consistently as perceptions of somatic anxiety. In similar research, Feltz (1988b) and Feltz and Mugno (1983) found that subjects' perceptions of autonomic arousal were better predictors of selfeefficacy that were physiological indices of arousal. Taken together, these findings suggest that one's cognitive appraisal of arousal may provide stronger efficacy information than arousal level per se. In conclusion. the results supported Hypothesis 2, and lent su;port to E: performance a of efficacy ; providing the In Hype: the stronges: supported th; coefficients rarginahy s; at least mar; waves_ Resu‘. “9‘39 7 somati '. 1935) Cf efficacy 1- atCEmpt :0 me 5.:- 4.. s. 5’ (2h;- . ‘u oer; L ‘04-‘1 anfie \- 80 support to Bandura's (1977, 1986) assertion that past performance accomplishments and emotional arousal are sources of efficacy information, with past performance attainments providing the strongest source of efficacy information. In Hypothesis 3, self-efficacy was hypothesized to be the strongest predictor of effort. Again, the data partially supported this hypothesis. The self-efficacy --> effort path coefficients were significant in Waves 1. 3. 5, and 8, and marginally significant in Waves 2 and 7. Thus, self-efficacy at least marginally predicted effort in six of the nine waves. Results of the full model analysis indicated that Wave 7 somatic and cognitive anxiety were the only other significant predictors of effort in the model. Bandura (1977, 1986) contends that judgments of efficacy are a major determinant of one’s effort expenditure. Research has supported this nocion in that individuals with a strong sense of efficacy have been shown to exert greater effort in an attempt to master a challenge, whereas those with weaker efficacy beliefs exert less effort when confronted with difficulties (Schunk, 1984; Weinberg, Gould, & Jackson, 1979). Similarly, self-efficacy was found to be a fairly consistent predic:or of effort in the present study The absence of stronger self-efficacy --> effort paths may be the result of players' biased perceptions. Perceptions of effort were assessed at the conclusion of each game, so players were aware of their hitting performance as they responded to the effort question. In situations where performance was poor, players' reports of effort expenditure may have been more an indication of the attributions they were making for their performance, rather than the actual amount of efi poorly while perceived as measure may ‘7 bias that at? self-presents. In Stun. hypOthesis t1". effort expend that higher p increased eff Moreover, the the dearth of 3pposed to pe I'C’r‘z'eratlily co Effort found Other than $6 variance in 6 also have bee ShOuld eXami; well as cOnS: influWe ef HYpothe 5," Roth Self‘ef Again, the r s: .JD'Qort for . 81 amount of effort they exerted while hitting. Performing poorly while exerting high levels of effort may have been perceived as indicative of low ability. Thus, the effort measure may have been tapping a self-serving attributional bias that athletes employed in attempt to make a positive self-presentation. In sum, the findings lend moderate support to the hypothesis that self-efficacy is a determinant of one's effort expenditure. Moreover, Bandura’s (1986) proposition that higher percepts of efficacy are associated with increased effort expenditure received partial support. Moreover, these findings are of particular relevance given the dearth of self-efficacy research employing effort, as opposed to performance, as a dependent variable. The moderately consistent relationship between self-efficacy and effort found in the present study suggests that variables other than self-efficacy may have accounted for some of the variance in effort. An inadequate assessment of effort may also have been responsible for the findings. Future research should examine alternative methods of assessing effort, as well as consider variables other than self-efficacy that may influence effort expenditure. 7 Hypothesis 4 stated that self—efficacy and effort would be the only predictors of hitting performance in the model, with self-efficacy being the stronger of the two predictors. Again, the results of the path analysis revealed partial support for this hypothesis. Effort was a marginal predictor of hitting performance in only one wave, whereas self- efficacy was found to predict objective performance in six of the nine waves. Conversely, effort was at least a marginally significant the nine wa‘-' found for £1”- pat . The abs paths in W8? concerns. 8* only when the performance efficacy will beyond one's performance. consisted of examination c performance 1' Performer's e the Opposing determinatio; ray throw ch. Control OVer hitting the l 82 significant predictor of subjective performance in seven of the nine waves, while only a trend toward significance was found for the Wave 3 self-efficacy --> subjective performance path. The absence of significant self-efficacy --> performance paths in Waves 4, 6, 7, and 9 was likely due to measurement concerns. Self-efficacy is purported to predict performance only when the execution of specific skills accounts for performance (Bandura, 1986, 1990). In other words, self- efficacy will not predict performance if factors that are beyond one's control are partially responsible for successful performance. In the present study, hitting performance consisted of the contact percentage of each player. A closer examination of contact percentage suggests that it is a performance index that is determined only in part by a performer's execution of specific skills. The performance of the opposing pitcher or the umpire also plays a role in the determination of performance success. For example, a pitcher may throw the ball extremely fast or may exhibit precise control over the location of his pitches, so as to make hitting the ball extremely difficult. A batter may execute the mechanics of his swing perfectly, and yet not contact the ball. Likewise, the umpire's decisions on called balls and strikes may have impacted the number of attempts players were permitted to make contact with the ball. Given this interdependence. self-efficacy may not have predicted performance because skill execution was not totally responsible for hitting performance. Another assessment issue stems from the fact that the performance measure was comprised of each players' average performance plate appear calculating bats. In cop efficacy mea; efficacy exp: If this were the game wouj Players' eff: stronger re may have bee. each plate a; efficacy duri D9rformance. in the Dreser field-baged r assessments t ComPEtltiQn. The til’m Efficacy QUE: influWed tl paths. Bandl H sEr formanCe é Deliefs may 1 n 131’: time period r For EXample' performance n a: . ~~flcaw' TIP 83 performance for a game. For example, if a player made four plate appearances, his performance was assessed by calculating the average contact percentage of those four at- bats. In contrast. each player completed only one self- efficacy measure per game. It is possible that players' efficacy expectations changed during the course of a game. If this were the case. the self-efficacy measure taken before the game would not have been an accurate assessment of players' efficacy appraisals while the game was being played. A stronger relationship between self—efficacy and performance may have been found if self-efficacy were assessed prior to each plate appearance. This way fluctuations in self- efficacy during competition could be assessed and compared to performance. Unfortunately, this protocol was not feasible in the present study due to situational constraints. Future field—based research should attempt multiple self-efficacy assessments to capture any changes in efficacy beliefs during competition. The time lapse between the administration of the efficacy questionnaire and actual performance may also have influenced the strength of the self-efficacy --> performance paths. Bandura (1978) has argued that if self-efficacy and performance are not measured closely in time, efficacy beliefs may be influenced by new experiences during the intervening period. Self-efficacy was assessed approximately 15-20 min prior to each game. Events occurring during this time period may have altered players' efficacy expectations. For example, heightened anxiety levels just prior to performance may have influenced players' percepts of efficacy. This explanation seems somewhat plausible, given the moderate q efficacy. . significant s€‘-effica< variables n< ' c eiperience- ' ‘. influentia- 7‘ The .e reciprocal However. t‘r efficacy-p. Studies (Fe 84 the moderate relationship found between anxiety and self- efficacy. The cognitive anxiety --> self-efficacy paths were significant in three waves, and somatic anxiety predicted self-efficacy in three other waves. It is also possible that variables not accounted for in the model, such as vicarious experiences and/or verbal persuasion, affected self—efficacy during this time period. Certainly, these potentially influential factors should be examined in future research. The results of the path analyses also denote the reciprocal relationship hypothesized to exist between self- efficacy and performance. In the hypothesized model, past performance was a significant predictor of self-efficacy, which in turn significantly predicted subsequent performance. However, this relationship was not as strong as the self— efficacy—performance relationship reported in previous studies (Feltz, 1982; Feltz & Mugno, 1983). Though contact percentage appeared to be an appropriate measure of hitting performance, as previously noted it still may have been influenced to some extent by external factors (i.e. umpire's calls; ability of pitcher) which may have inhibited stronger associations between self-efficacy and performance. The finding that effort positively and consistently predicted subjective performance supports the contention that players’ reports of effort may have been tapping their attributions for performance. Better subjective performances were associated with higher percepts of effort, whereas poorer perceived performances were associated with less effort expenditure. In both instances. attributing performance, at least in part, to effort expenditure could have been a self-serving bias on the part of the players. For example effort expei about the pE ‘ his COHCI‘OI. attributed t Still exhib" through inc: This no that subject Efficacy bel 85 For example, attributing a good performance to heightened effort expenditure suggests that a player was able to bring about the performance through effort, a factor that is within his control. Similarly. poor performances could have been attributed to a lack of effort, in which case the athlete still exhibited control, and could improve performance through increased effort. This notion is supported by McAuley (1990), who found that subjects making internal but unstable attributions for dropping out of exercise programs had higher efficacy expectations with regard to their ability to overcome possible barriers in future exercise programs. In the present study, athletes may have continued to exhibit high efficacy beliefs due to the attributions they were making for performance. It should be noted that though greater effort expenditure has been shown to be associated with lower perceived ability (Nicholls, 1978, 1980). heightened effort is typically perceived as a socially desirable behavior within a sport team framework, and therefore may not have been indicative of incompetence. The fifth hypothesis stated that subjective measures of performance would be stronger predictors of self-efficacy than would objective measures of performance. This hypothesis was based on Bandura's (1986) assertion that individual’s efficacy expectations are influenced more by how they interpret their performance successes and failures. than by performance attainments per se. Thus, the predictive capability of perceived performance was compared with that of actual performance. Virtually no support was found for the influence of subjective performance on self-efficacy beliefs. as subjectiy and subje ti qaestionnair were very di their hittir performance efficacy bej hive also f: COHCEDtuali; Haney (1991' a measure 0: SatisfaCtiO] 1 '7 “3‘61 of lie; that self-e. .Gerforman c e pe.forme* emeria for subjectlfe I is an area SLppOrt .idS the we“ ‘ a“; j «'1 'w “new inf 86 as subjective performance predicted self-efficacy in only one wave, whereas objective performance predicted efficacy in six of the nine waves. This finding may be the result of incongruous efficacy and subjective performance measures. The hitting efficacy questionnaire may have been assessing aspects of hitting that were very different from the criteria players used to rate their hitting performance. Thus, beliefs about hitting performance may have had little relevance to the hitting efficacy being assessed. Haney (1991) and McAuley (1985) have also found subjective assessments of performance to be weak predictors of self-efficacy. It should be noted, however, that the lack of significant perceived performance — —> self-efficacy paths may also be due to poor conceptualizations of perceived performance. For example, Haney (1991) used subjects' satisfaction with performance as a measure of perceived performance. Clearly, one's satisfaction with performance does not necessarily reflect level of performance attainment. Thus. it is not surprising that self-efficacy was not found to predict perceived yperformance in that study. Perhaps a more appropriate assessment of perceived performance may be obtained by having subjects estimate how well they performed relative to other 'performers. In this way, subjects would use a common criteria for subjective assessment. Certainly, the use of sdbjective performance assessments in self—efficacy research is an area that needs further attention. Presently, little support has been found for Bandura's (1986) assertion that the way in which one interprets performance provides stronger efficacy information than performance attainments per se. PA! 1V (‘nc The re earlier cat; efficacy wa performance 1934). How. of performa: performance :snsistently path coeffic 1388b) and E ;erformance influence of :ime, The fin ;a:h analy: i fOLlIld betWee lated w indies of m Skill Unde . named Vir' 87 Conclusions The results of the path analysis were consistent with earlier causal studies in motor performance in that self- efficacy was found to be both a cause and effect of performance (Feltz, 1982, 1988b; Feltz & Mugno, 1983, Schunk, 1984). However, self—efficacy tended to be more of an effect of performance, rather than a cause of performance. The past performance --> self-efficacy path coefficients were consistently stronger than the self-efficacy --> performance path coefficients (see Table 14). Similarly, Feltz (1982, 1988b) and Feltz and Mugno (1983) found that the influence of performance on self—efficacy increased over time, while the influence of self-efficacy on performance decreased over time. The findings in the present study differ from previous path analytic studies in that the reciprocal relationship found between self-efficacy and performance in the present study explained only a small portion of the variance, whereas previous research has found that self-efficacy accounted for a large portion of the variance in performance (e.g. Feltz, 1982; Feltz & Mugno, 1983). In addition, previous studies have consistently reported a direct, significant path from past performance to subsequent performance. A direct past performance —-> performance path was not found in the present study. These differences may be due to contextual factors associated with performance. For example, in previous causal studies of motor performance subjects performed a closed skill under controlled conditions. That is, the environment remained virtually unchanged for all performance trials. thus. it is and perform preSent stuC Variable: d3 a throw! :3 itch. 1 ;;:cherS 95' p;:chers “1 ;Layed on a under varia almost unpr influential that prior performance performance variance ir remarkable be fairly c riable cc The f: .5: 88 Thus, it is not surprising that one's efficacy expectations and performance remained somewhat stable over trials. In the present study, players performed an open skill under variable, dynamic conditions. For example, players had to hit a thrown ball, whose speed and location varied from pitch to pitch. Moreover, players had to adjust to different pitchers each game, and often times faced a variety of pitchers within the same game. In addition, games were played on a number of fields, against a variety of opponents, under variable weather conditions. Given the changing, almost unpredictable environment and the potential for the influential effects of other factors, it is not surprising that prior performance did not predict subsequent performance. Nor is it surprising that self-efficacy and performance accounted for only a small portion of the variance in the data. On the other hand, it is somewhat remarkable that self-efficacy and performance were found to be fairly consistent predictors of each other under such variable conditions. The findings from the path analysis also suggest that self-efficacy and performance are not only reciprocally related, but also that self-efficacy acts a cognitive mediator of performance.' According to Baron and Kenny (1986), mediation would have occurred if past performance and/or anxiety exerted a direct effect on self-efficacy, which in turn directly affected effort and/or hitting performance. An examination of the hypothesized and full :models indicated that self-efficacy was the strongest and most.consistent predictor of effort and performance in either rmmdel. Moreover, the effect of past performance on sasequent 2' 3! 5' an mediated by efficacy me Waves 2, 3. waves 1, 2. The so variable co efficacy th oozed, mos: elements in PCpulations conditions Studies ha\ sexings, k 89 subsequent performance was mediated by selfeefficacy in Waves 2, 3, S, and 6, and the effect of anxiety on performance was mediated by self-efficacy in Waves 1, 2, 3, 6, and 8 . Self- efficacy mediated the effect of past performance on effort in Waves 2, 3, 5, and 7, and the effect of anxiety on effort in Waves 1, 2, 3, 7, and 8. The support found for Bandura’s (1977) model under such variable conditions also speaks to the utility of self- efficacy theory in actual sport settings. As previously noted, most of the previous research examining the causal elements in Bandura's model have employed nonathletic populations engaging in novel tasks under controlled conditions (e.g. Feltz, 1982, 1988b; McAuley, 1985). Other studies have examined athletes performing in competitive settings, but examined only the correlational relationships between self-efficacy and performance (e.g. Barling & Abel, 1983; Lee, 1982). Still other studies have undertaken causal investigations of athletes performing in contrived situations (Fitzsimmons et al., 1991; Haney, 1991). The present study examined the causal relationships hypothesized in Bandura's model, using athletes competing in their chosen sport, and replicated the causal and mediational effects of self- efficacy on performance found in controlled settings. Field- based support for the causal mechanisms of self-efficacy theory had been missing in the literature relating to sport and motor performance prior to the present study. . J 1. . The finding that self-efficacy significantly and consistently predicted effort and performance has strong 9O implications for sport psychologists, coaches, and teachers. If self-efficacy acts as a mediator of effort and performance, then factors that influence self-efficacy may be manipulated to facilitate performance and increase effort toward performance. For example, sport psychology consultants may provide athletes with cognitive strategies that enhance performance through the persuasory effects they exert on self-efficacy. Positive self-talk, goal-setting, and imagery techniques are three performance enhancement strategies that may operate through the effects they exert on efficacy beliefs. Likewise, coaches may structure practice situations to maximize mastery experiences, which in turn should enhance efficacy expectations and facilitate future performance. For example, practices may be structured so that players successfully perform various parts of a skill or gradated versions of a skill prior to attempting the whole skill. Players' mastery of the components of a skill should enhance efficacy expectations toward performance of the entire skill. Conversely, placing inexperienced players prematurely in a taxing situation where they are likely to fail will undermine their sense of efficacy, and make it difficult for them to master the skills necessary to successfully compete in that situation. Teachers and coaches may also facilitate future performance by helping students/athletes focus on salient features of previous performance. In the present study, athlete's subjective performance assessments were not predictive of specific measures of self-efficacy. It is likely that athletes were focusing on aspects of hitting 91 other than making contact with the ball, and thus a poor relationship existed. On the other hand, actual contact percentage was a strong predictor of "contact" efficacy. Perhaps if players had been instructed to subjectively rate their hitting performance in terms of making contact with the ball, a stronger relationship would have been found between subjective performance and self—efficacy. Thus, by directing students'/athletes' attention toward specific aspects of performance, percepts of efficacy may be enhanced, leading to stronger performances. In addition to athletic settings, the work of Schunk and colleagues (Schunk, 1984; Schunk & Cox, 1986; Schunk & Gunn, 1986) provide support for the enhancement of academic performance through the manipulation of self—efficacy beliefs. Attributional feedback related to effort and ability has been shown to enhance childrens’ self-efficacy beliefs and academic performance. Similar feedback strategies could be employed in classroom and athletic settings in an effort to raise percepts of efficacy, which may lead to greater effort and improved performance. . .r . E I] S! 1 Though the present study found support for Bandura's (1977, 1986) model of self-efficacy, the results should be interpreted with caution due to several methodological limitations. First, the sample size was relatively small. Only fifty-three baseball players participated in the study. Because not all of the players participated in every game, only a portion of the total sample (72%) was employed in the path analysis at any given time. Pedhazur (1982) suggests 92 that a sample size of at least 100 is necessary to conduct path analysis. However, most other path analytic studies of self-efficacy in the literature pertaining to sport and motor performance have employed smaller sample sizes (e.g. Feltz, 1982, 1988b; Feltz & Mugno, 1983; McAuley, 1985). Thus, the sample size employed in the present study is not uncommon in research relating to sport and motor performance. In addition, the small sample size may have influenced the g? goodness-of-fit tests. However, the use of the Q ratio added confidence to the accuracy of the 32 tests because it was not affected by sample size. Therefore, the potential problems associated with a small sample may have been reduced to some extent. Second, as previously noted, the perceived competitive anxiety measure used in the present study was administered retrospectively, rather than prior to performance. Players were asked to recall how anxious they felt just prior to hitting. This protocol was implemented on the recommendation of Feltz and Mugno (1983) in an effort to assess perceptions of anxiety as close as possible to actual performance. It is possible that players’ recollections of perceived anxiety were inaccurate, or that players' appraisals of anxiety were influenced by their hitting performance. Similarly, the perceived effort measure was completed following performance and may have reflected a self-serving attributional bias to account forperformance. Thus, the relationships found between these variables may not be an accurate portrayal of their association. Finally, the influence of extraneous factors (e.g. 93 opposing pitcher) on self-efficacy perceptions and performance was not controlled. Thus, the extent to which environmental and/or situational variables impacted efficacy expectations and performance is not known. Certainly, the lack of significant path coefficients for the various hypotheses, as well as the small amount of variance accounted for in the data may be attributable to factors outside of the model. . . The results of the present field study provide support for Bandura's (1977) model of self-efficacy in an actual sport setting. The causal relationships found in previous controlled research settings were replicated in a field setting, though perhaps not as convincingly. Nonetheless, the results of the present study provide ecological validity for Bandura's model. Contrary to previous research (Feltz 1982, 1988b; Feltz & Mugno, 1983), past performance accomplishments did not exert a strong, direct influence on future performance. Instead, in support of Bandura's (1986) contentions, the influence of past performance was mediated by self-efficacy. The discrepancies in the findings may reside in the type of tasks employed in the various studies, or the conditions under which subjects performed the tasks. Past performance exerted a strong direct influence on future performance in controlled settings using a closed skill, but did not influence performance in a field setting employing an open skill. Future research should investigate potential task types and contextual factors that may play a role in the past 94 performance - future performance relationship. Perhaps the predictability of one's performance meditates the extent to which past performance predicts future performance. Future research should also focus on the development and use of self-efficacy measures. Typically, researchers have developed their own measures of self-efficacy in order to address a specific question. Researchers have asked similar questions, but have employed vastly different measures of self-efficacy in an attempt to address these questions. The development of valid, reliable measures of self-efficacy may be facilitated by repeated use of efficacy measures across a variety of populations and research settings. The work of Feltz and associates (Feltz, 1982, 1988b; Feltz & Mugno, 1983; George et al., 1992) illustrates the repeated use of efficacy questionnaires across a number of studies. Future research should employ similar strategies and attempt to replicate the findings of self-efficacy studies in both controlled and field settings by employing similar measures of efficacy and performance. Until this type of research is undertaken, the ways in which self-efficacy and performance are assessed will likely determine the strength of the relationship between self—efficacy and performance. Another area in which future research should focus pertains to the sources of efficacy information. Typically. research has examined the effects of only one or two sources of efficacy information on self-efficacy beliefs. Most of the causal examinations of the effects of efficacy information on self—efficacy have employed past performance éand anxiety as the only sources of efficacy information. Future research should also investigate the causal 9S relationships in the self-efficacy model using vicarious experiences and verbal persuasion. The causal studies that have employed these two sources of efficacy information have generally provided support for Bandura's theory (e.g. Feltz & Riessinger, 1990; Fitzsimmons et al., 1991; McAuley, 1985). In addition, future research should attempt to examine all four sources of efficacy information simultaneously. Such research would provide valuable information relating to the most salient sources of efficacy information. Moreover, this research may provide some indication as to whether the four sources of efficacy information are accounting for unique variance in self-efficacy or sharing the variance they explain. As noted in Chapters 1 and 2, self-efficacy is hypothesized to influence behavior in terms of choice of activities, effort expenditure, persistence, thought patterns, and emotional reactions. However, much of the self- efficacy research has examined the effects of self—efficacy on performance. Though support has been found for the self— efficacy - performance relationship and research should continue to examine this relationship, future research should also focus on the behaviors originally proposed by Bandura's (1977) theory. The present study attempted to assess the effects of self-efficacy on perceived effort expenditure. Perhaps a physiological assessment or other objective measure of effort is required to address this relationship. JNOnetheless. future research should continue to investigate the influence of self-efficacy on behaviors such as effort. persistence, choice of activities, and thought patterns. Finally, the results of the present study supported the 96 hypothesized causal mechanisms in Bandura's model of self— efficacy using experienced athletes competing in an actual sport setting. Several other research endeavors have also supported the relationships in self-efficacy theory using experienced athletes (Barling & Abel, 1983; Fitzsimmons et al., 1991; Haney, 1991, Lee, 1982). However, these studies have either used noninferential techniques to examine athletes competing in actual sport settings, or causal techniques to investigate athletes performing in contrived situations. Future research should continue to engage in field-based, causal investigations that examine athletes competing in their chosen sport. This line of research would delineate further the predictive generality of self-efficacy across sports, and provide an indication of the utility of self—efficacy theory as a conceptual framework from which to examine self—confidence in athletic pursuits. APPENDIX A APPENDIX A Hitting Self-efficacy Scale PLEASE INDICATE HOW CERTAIN YOU ARE OF YOUR ABILITY TO PERFORM THE FOLLOWING HITTING TASKS IN TODAY'S GAME AGAINST BY PLACING AN '1' O 10 very uncertain 20 30 IN THE APPROPRIATE BOX. 40 60 70 80 very comm Putthehnllinplnyltimein4nt-hats Putthehnilinplnyltimein4nt-hnts Pntthehnllinplny3timein4u-bnts Putthehnlllnpiny4timein4nt-bats Drive in a runner from 2nd base 1 time in 4 tit-bats Drive in a runner from 2nd base 2 times in 4 at-bats Drive in n runner from 2nd base 3 times in 4 nit-hats ' Drive in n runner from 2nd base 4 times in 4 alt-bats Lay down I sacrifice hunt in the 1st, 2nd, or 3rd inning Lay down n sacrifice hunt in the Inst inning Kit I sacrifice fly in the 1st, 2nd, or swim-15 mtnncriflcenyinthelnninning Execute n hit-nnd-run in the let. 2nd, or 3rd inning Enecute n hit-nad-run in the int inning Drive in the winning run m 04 outs in the Inst inning Drive in the winning run mi 2 outs in the inst inning Lay down n suicide squeeu hunt in the enrly innings of a tie gnlne Lny down n suicide squeeu hunt in the inst inniwtie gnme 97 APPENDIX B APPENDIX B Competitive Anxiety Scale Directions: A number of statements which athletes have used to describe their feelings about competition are given below. Read each statement and then circle the appropriate number to the right of the Statement to indicate how you felt just before hitting. There are no right or wrong answers. Do not spend too much time on any one statement. but choose the answer which described your feelings just before hitting. Not at Moderame Very Much All Somewhat So So 1. I was concerned about hitting in today's game 1 2 3 4 2. I felt nervous about hitting. l 2 3 4 3. I had self-doubts about hitting 1 2 3 4 4. I felt jittery at the plate. 1 2 3 4 5. I was concerned thatI may not hit as well in today ’5 game as Icould 1 2 3 4 6. My body felt tense at the plate. 1 2 3 4 7. I was concerned about striking out. 1 2 3 4 8. I felt tense in my stomach. I 2 3 4 9. I was concerned about choking under pressure. 1 2 3 4 10- My body felt relaxed at the plate. 1 2 3 4 11. I was concerned about hitting poorly 1 2 3 4 12' My heart was racing while standing 1 2 3 4 at the plate. 13- I Was worried about reaching my hitting goal 1 2 3 4 14- I felt my stomach sinking as I stood at 1 2 3 4 the Plate. 15' I was concerned that others would be disappointed with my hitting performance. 1 2 3 4 16' My hands were clammy. l 2 3 4 17.1 Was concemedl wouldn't be able to concentrate while I was hitting. 1 2 3 4 18- My body felt tight while standing 1 2 3 4 at ““3 plate. 98 APPENDIX C APPENDIX C Subjective Performance Scale and Effort Scale [:3 PllICE nu '3' IN "I! son It vou mu nor Ptnv IN ronnv's can: HITI'ING PEBFOBMflNCE HHTING Please circle a number on the following scale which best indicates HOW WELL YOU THINK YOU HIT IN TODAY'S GAME. extremely poor moderate good extremely 139°r well I I | | | I If I l | l l 1 2 3 4 5 6 7 8 9 10 Please circle a number on the following scale which best indicates HOW MUCH EFFORT YOU PUT INTO HITTING IN TODAY'S GAME(S). LIIIIIIIIII lllllllllll 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 99 APPENDIX D APPENDIX D Informed Consent Form Michigan State University Department of Health and Physical Education Investigator: Thomas R. George hereby agree to participate as a volunteer in a scientific I, study of self-confidence and athletic performance as an authorized part of the research program in the Department of Physical Education at Michigan State University under the supervision of Dr. Deborah Feltz. The purpose of this study is to examine the relationship between self-confidence and baseball hitting performance. In addition, this study will examine the effects of anxiety on self-confidence and performance. You will be required to complete a battery of questionnaires assessing self-confidence, anxiety, and performance before and after nme successive baseball games. You will remain anonymous and your responses will be known only to the investigator. The study and my part in the study have been defined and fully explained to me and I understand this explanation. I have been given an opportunity to ask whatever questions I may have had and all such questions and inquiries have been answered to my satisfaction. I understand that my participation in this study does not guarantee any beneficial results to me. I understand that any data or answers to questions will remain confidential with regard to my identity. Within these restrictions. results of the study will be made available to me at my request I FURTHER UNDERSTAND THAT I WILL NOT BE PEN ALIZED IN ANY WAY FOR REFU SING TO PARTICIPATE IN THIS STUDY, AND THAT I AM FREE TO WITHDRAW MY CONSENT AND DISCONTINUE MY PARTICIPATION AT ANYTIME. SOCIAL SECURITY NUMBER DATE DATE OF BIRTH SUBJECT ’S SIGNATURE I the undersigned, have defined and fully explained the study to the above subject. TDATE INVESTIGATOR'S SIGNATURE lOO APPENDIX E APPENDIX E Table 1. Means and Standard Deviations for Variables. Som Cog Self- Hitting Wave Anx Anx Efficacy Effort Performance SPR Past Performance College M .786 5.60 SD .238 1.63 High School M .867 4.96 SD .291 1.90 wave 1 College M 13.22 18. 7 86.77 8.89 .807 5.25 SD 3.1' 5.97 13.62 1.37 .190 1.77 High School M 13.89 18.50 73.54 8.06 .816 5.16 SD 3.85 3.75 16.65 1.12 .254 2.04 wave 2 College M 12.57 15.90 92.00 8.85 .841 5.46 SD 2.98 3.97 9.19 1.09 .223 2.34 High School M 14.61 17.94 77.50 7.06 .843 4.44 SD 6.20 6.01 17.75 2.19 .223 2.82 wave 3 College M 11 73 15.64 91.90 9.09 .890 5.58 SD 2.90 4.21 10.16 1.04 .189 2.39 High School M 13.67 17.22 78.03 8.00 .858 5.89 SD 6.01 5.47 16.02 1.63 .273 2.49 wave 4 College M 12 15.91 91.70 8.89 .908 5.50 SD 3 ‘6 3.62 12.01 1.41 .161 2.90 High School M 14.00 16.30 83.07 8.47 .850 6.25 SD 5.08 5.89 11.95 1.74 .270 2.24 Som anx = somatic anxiety Cog anx = cognitive anxiety SPR = subjective performance rating 101 102 Table 1 (Cont’d). Som Cog Self— Hitting Wave Anx Anx Efficacy Effort Performance SPR wave 5 College M 12.04 15.78 92.71 8.73 .871 5.17 SD 3.78 5.03 9.18 1.78 .225 2.74 High School M 15.21 17.53 82.02 8.40 .748 4.63 SD 4.37 3.94 12.91 1.72 .215 1.89 wave 6 College M 11.13 14.52 94.57 9.39 .867 5.74 SD 2.72 4.55 6.60 0.89 .135 2.38 High School M 13.89 17.11 78.68 8.92 .838 5.86 SD 5.99 7.16 16.63 1.38 .278 2.94 wave 7 College M 12.71 16.24 92.16 9.41 .757 4.74 SD 2.62 5.90 11.58 0.94 .223 2.52 High School M 15.00 16.71 79.29 8.16 .718 5.29 SD 4.78 5.61 18.20 2.12 .339 2.72 wave 8 College M 12.50 15.28 93.12 9.24 .886 5.22 SD 2.85 4.34 8.79 1.09 .164 1.77 High School M 14.85 17.65 83.75 7.42 .721 5.20 SD 5.23 7.24 12.65 2.22 .338 2.33 wave 9 College M 12.12 14.94 93.04 8.73 .847 3.89 SD 3.57 4.66 9.44 1.35 .207 2.47 High School M 12.61 14.56 78.87 7.89 .847 6.33 SD 3.55 5.96 21.33 2.17 .236 2.52 APPENDIX F APPENDIX F Intracorrelations for Independent Variables r 1 1.1L mocmeuouuma mcfluufln msofl>mua u mum HO. V Q .1 mo. v Q + oo.H OH.- ma. vm. 8H. +mm. ma. +mv. ma. m oo.H Ho. mm. mo. «a. No.1 mo. Ho. m oo.H +nm. mo.- mo. mm. 00. Ha. ma.u b oo.H mo.- mo. Ho. ho. ma. mo.- m oo.H mm. saw. em. +ov. ma. m oo.H +mm. «Hm. om. ma. 6 oo.H 6m. mm. em. m oo.H +ov. ma. m oo.H NH.- H m>m3 m m o m 6 m m H mum o>mz .maa mm>oz u mocoEMOMMDQ mcfluuflz we mcoflumenuououucH .N mHDGE 103 104 Hoo. v d « oo.H 4H5. sow. «Ho. sow. «we. *Nm. «mm. «vm. m oo.H 4mm. 4mm. «mm. amm. «mb. sob. *mb. m oo.H «hm. sob. sow. «mm. 4mm. «vb. b oo.H «mm. «mu. «mu. *vm. “am. 0 _oo.H 4mm. «Mb. «mm. «em. m oo.H «vm. saw. «mb. v oo.a «mm. 4mm. m oo.H «mm. N oo.H. H QHMZ m m b o m w m N H m>o3 .m-a mo>mz u >omoflwmmawamm wo mcoflbmawuuoomuucH .m manna 105 Ho. v a + mo.va* oo.H +ob. +mm. +mm. «mv. +mm. om. +mv. +Hm. m oo.H +mb. +mb. +mw. +>b. +06. +Hm. +Hm. w oo.a +om. +mv. +mm. +mm. +mm. +mm. b oo.H +mb. +Hw. +mb. +Hm. +bm. m oo.H +vm. +mm. +mv. «mm. m oo.H +mm. +66. +mm. v oo.a +mb. +bh. m oo.H +mm. m oo.H H “Hag m m b m m w m m H w>m3 .m-H mm>mz - znmflxc< m>fluncmoo no mcoflumamuuoumuncH .e magma 106 Ho. v a + mo. v a + oo.H +vm. +06. +mm. mo. +mm. Hm. bH. NH.: m oo.H +om. +vv. mm. +Hm. +om. mm. +om. m oo.H +mw. +mv. +mm. +wv. vm. om. N. oo.H +wm. +mm. +mm. «m. mm. m oo.H Hm. Hm. 0H. MH. m oo.H +mm. «ow. +mv. v oo.H +ov. +ow. m oo.H +vv. m oo.H H ufimz m m b o m w m N H w>m3 .muH mw>m3 : muwac< oHumEom mo mcoHumHmuuoomuucH .m mHQme 107 Ho. v d + mo.va+ oo.H +mm. +mb. mo. «bv. +mm. lgh. mm. mm. m oo.H «vv. HH. HH. +mb. +mb. +hm. «mv. m oo.H +mv. mm. +mm. +mm. +mv. +om. b oo.H mo. mo. mm. +nv. vo. m oo.H «mm. «mm. mm. +Hm. m oo.H +mm. +mm. vH. v oo.H +mv. +mm. m oo.H +mm. m oo.H H quz m m h m m w m m H ®>63 .mIH mm>mz I DHOuwm m0 mCOHumHmHHOUMHuCH .w mHDMB APPENDIX G APPENDIX G r11 . lii'fllfii‘ lJ-ZINH I orljt HUMUHmwm-MHmm H mm humecm UHDMEom u sow humecm m>HuHcmoo u 000 mmcHumu mocmEHOHHmQ m>Huowmnsm u mam mocmEHOHHmQ mcHuan msoH>mHQ H mm H0. v a .. 00. v a . 0H. 0H.- 0H. .00.- 00. 0N. 00.- 0H.- N0. uHOHHm-mmm N0. 00. 0N. .N0. H0. NH. 0H. .00. ..00. mm-mmm 00. NH. .00.- 00. 0H.- 0N. H0.- 0H.- 0H.- sow-mam 0N. 00. ..00.- NH. 0H. 0H. .00.- 0H.- 0H.- moo-mam HH. 0H.- 00.- NH.- 00.- 0H. 00.- 00.- 00. 000000-00 ..00. 00. ..00. ..N0. ..00. 0N. .00. ..00. 0H. 00-00 0H. 0H. 0H.- 00.- 00.. 00. 0H.- 0H. 0N.- sow-00 0N. HH.- .00.- H0.- 0H.- 00. 00.- 00.- 00.- 0oo-00 .H0. .00. ..00. ..00. 0N. 0H. 00. HN. .0N. mam-00 moHQmHHm> 0 0 0 0 0 0 0 N H w>mvz .w>m3 >9 mmHQmHHm> uchconccH mao€< mcoHumHmHHou .0 mHnme 108 109 Ho. v Q .. mo. v Q + mm. +«Hv. mm. HN. «mm. HH. «hm. Hm. ++vv. uHmemnmm $0.: MH. mo.n ++mv.| Hm.- OH. hm.u Ho.| mH.u uncummuoou «.mv.n .mm. «+mv.n «mm.| 0N.: ««Hv.u «*Hm.n +mm.u «mm.n mmumOU bH.u arm. «mm.: mm.- NN.- mo. +mm.u mo. Hm.| DHOwwm-Eom +bm.n .«mv. «+vm.n mm.| 0N.: 0N.: «+mv.n Ho.| «aom.n mm-Eom ++Nm. 4+mh. +¥ho. ++ob. «+mw. «.mw. «+0b. «amm. «.mv. mou-Eom moHQmHHm> m w h m m w m m H m>m3 .AU.DCOUV b mHQME APPENDIX H APPENDIX H FIN .11; viauinug 5&1 0‘ A» s .01. 1i mucoEHOMHmQ ucHuan u uHm HomuHuum-uHmm H mm humesu oHuoEOm u xsoeom Huoncw 0>HuHcmou u Mammoo wUGQEHOHNwQ 0>Huummnsm u mam mUCMEMOMHmQ mcHuqu mon>oHQ H mm Ho. V Q t« mo. V Q t 0H. v Q + +Nvm. +Nmm. +NHm. mvN. «mv<. +vom. «comm. mvN. comm. mmm All UHOMNW HbN.I mmo.l OHo.I mwo. Hmo. mHo.I +NmN. bNo. moo. mmm All mm vwm. OHN. be.I ovo. +me. NVH. boo. moo.l mod. HM: All UHOHMN MHH. smvm. Hmo. +omN. «mom. mNN. +on. +mmm. «mflfi. was All mm bNN. avHv. +mbm. oHN. «mam. wOH. :th. +MHm. «cmmv. UHOuum A1: mm OHo. mmo. mmm.l mmH. mmo.| PMH. Nmo.l +mmN. amom. mm All Mmm «#wa. «PH. ttmmv. «tva. iObv. OHN. #mvm. «ivmv. v¢o.t mm All mm NmH.I «mom.l cHhv.| moo. omo. Hwo. mmH.l NHN. cvmv.n Wm All Xflflfiom ONH.I mNH. hOH.I «Hmm.| mmH.l smmv.l +mmm.l ammv.l bmo.l mm Al! xGCmOU comm.| ooo. +mmm.| mvo. Nbo.l NMN. +HHm.I MHN.I NOH.I XG03 .m>m3 >3 Hwooz omNHmwSDOQmm mnu 00m muGoHUHuuooo comm .wH mHnme '110 111 r i .ql 1.11.4. £0151»! 0.1.u4. _. .. 51%|..V‘L ll 1| f|-- II mmcHumu mUGMEMOuqu 0>Huumn33m mam mucoEMOMHoQ mnHuan uHm humecm oHunEOm u xcnaom humeco 0>HuHcmoo u Mammou wUQMEHOMHmQ mcHuan mon>0uQ H mm Ho. V Q «o 00. v a . 0H. v Q + 00H. 00H. N00. H0N.- 00N. 0HH. 000. H00. 000. UH: A- 000000 000. 00N. 000.- +000. NH0. 000. 0HN. 0HN. .000., 0H: A- mm 000.- 000.- HN0.- 0H0.- 000 - H0H.- 00N. 000. .0N0. 0H0 A- xzM3 .w>mz >9 Hmooz m>Hmusowm HHHDQ mnu Mew mucwHonumoo zoom .mH mHQMB 112 Ho. V0.0: mo. V Q t 0H. v Q + «Hhv. +vmm. +mmm. mmH. +vvm. mmm. «cmmm. «va. aNHv. QO All UHOuum +mmm.a oom.1 mwN.I +vom.| QMH. mfiH.I mmm.n th.I mOH.I QO All mm mmm.u va.n bov.l ¢«mbh.u mNH. me.I omo.l NMH.I omN. QO Au: xcdmou bmH. hmH.I Nmo. QMH. Hmm.u moo. NHo.I +omm.l ovN.I QO All Ndéfiom cmmv. 0HN. vHo. a¢omv. mom. «mam. boo. mvm. HmH. QO AI: QO Hmo. mmo.t mmH. Hoo. NmH.I moo. NNo. m¢H. mmo.l QO AI: QQ SDMQ m m b m m v m N H m>oz .Av.ucouo mH mHQme APPENDIX I Mutt APPENDIX I .cmme zoom BonQ mommcucoHMQ CH cm>Hm mum mcoHumH>mo UnmoGMDm "muhée Amw.mHv AOH.mHV Awo.mH. AOH.bHo AbH.mHo Hmm.mHv Hon.mHo .Amm.HNV Afim.NN0 mmefiuv :HmnomflgocN mm.mm mb.vw mm.Hw mH.mm oo.mw mm.mm mm.ow mm.mm NH.wv sofiuuocceunfi m>fiXH Amo.mHv Hmm.mHv Abo.mHv va.vHv Hoo.mHv HmH.mHo Amm.bHo AmH.omo Abb.HN. mosfiuv :HNumeHocN mm.vb hm.vh mo.mb Hb.mh NN.vb b¢.¢b ma.mb mm.mw mm.mm eofluhoccpufiz m>fixu AbH.mHV Amh.mHv Amm.mHV Amp.MHv Amm.NHv Hmw.mHV Amm.vHo Amv.mHo Hmv.mH. mmefluv cHHmeQHvCN mv.vm 2.00m mme mm.mm mb.mm >¢.vm mmém mem H082. .003 .855» 5 003.8 HHm.vmo Avo.Hmo HbH.NmV Hwo.Hmv Am¢.omv Hmm.omv Amm.mmv Amm.mmv HHN.bm. mwsfiuv 900 mo.mb vo.mh Hm.mb mm.mb mm.mb mm.mh mm.mb mm.mb mv.Hm xmducH difi Qfiyuam Hmb.mHv Hmm.vHv Hob.bHv ANH.mHv Abm.vHv AbN.wH AvH.mHo Abo.mHo Amm.mmv meHuq.cH 0 mm.mm hm.mw mH.vm vm.mw HH.mm HH.mm mH.mm mm.mm No.mb >020cH ofifi Qfiuusm Amm.mHo Arm.mv Hmv.mHv Hmm.mHo Amo.0Hv HNH.OHV Amw.mHv va.MHo HwH.vHv mmefiuv cHN ov.Hm or.mm oo.om mH.Hm bw.mm mm.mm vH.Hm vv.om mm.bm >010c0.26n Qfiuuzm Hmo.qHV Hmm.mv HHm.mHv Hbm.OHV HvH.wv Aww.bv Hom.hv Hbm.bo HHm.mo mwsfiuv cHH Nv.vm mm.mm mv.mm mn.vm mm.mm mm.wm mv.mm HH.mm mm.mm >mfiuc0 H12.wcu HE .HOu >uHHHno “50% :H 30> who cHouumu 30x. :oHummDO m m b m m 0 m N H o>53 .mm>m3 mmouo< mEmDH Homonwm-mem HmsoH>HocH Ho“ mcoHuoH>oQ oHMUCMDm ocm memo: .wH anm9 113 114 F's-lllgqg P1... .. 1.3.13” Gmémo 30005: ANXHNV £20.63 Ammo: 3mm: 26.003 301m: 390.: 9:9: 803 0: mo.mb FEE. 222. 22:. 05.3. 002:. 0H.H~. oo.m~. mm.vm can-DE musomxm :35: Hom.mHv 2:03 $0.03 AHHJHV HmHéHv 3m.va 23.93 3N6: mchcH 0H0 HEN ..umH 0.0.2. m0.om 06.3. 0H.mm mm.om Hm.m\. 30.2. 22:. mo.m~. c0 55-an wnsomxm Hmm.mmv AvH.mHV AHo.mHV Hmm.mHv Avm.va Hmb.mHv HHo.omv Amb.bHv Amm.mHV mash: uflflncH ov.:. mm.mb mv.m.\- mm.m0. mm.mb mm.m0. 092. mm.mb mpo 0H0 333000 0H: :0..HNV 30.0: Ammé: 3106: 3.0.03 2.0.93 3.70.: 3min: :06: 655:. on... .ocN 3:: :32. ENE. mm.om mm.mm mmHm mm.mb than HH.m\. 2.3; 5 0H0 3:20.00 0H: 32.03 30.0: Ammo: 30.va 3.6.0: 3H0: 26.0.: 30.2; Smog 0.30:: 00.3 5 0:3 mw.Hm 0m.vm mv.vm om.mm vv.mm mH.mm mm.vm mm.mm NH.mm worznofiwczzc>sH 9.1:; Homo: :v.mHV 30010.: HmmHHV 2.de Ammo: AmméHo 3m.m3 arm: 20 65 .03 cH 055 Siva oo.om thaw mva vv.mm mo.mm Emma wm.mm vm.mm 8303.00 Eon >3 :mHm. Ammomv HN.-0.2.; $0.6: HHmSNV Hmoémo :«émv Smog SEN”: mos: 0 cH 0 00.00. HEN mH.Hm mv.mm hb.mv mb.vm mm.om vm.om mm.mv mm.mv mb.¢m sofiuuwfihfi SHw>fix0 .HOu huHHHnm Hzox cH 30> mum :Hmuumu 30:. :OHummDO a n H. o 0.. H0 n V H w>03 .HU.uCOUo mH GHQMB 115 memo HmH.mmo HHm.omo Hmv.Hmo va.omv HHo.HmV Hmv.mmv Amm.omo Amb.mHv AHm.omv mHuwo QbécHufimH mm.mb 0.00.2. mmob mode omit. main-.0. wmob HH.m0. 2.4:. Haas ononm 5.600 0.3 0500 Avm.mHv AHH.mHV Hmo.>Hv HHH.mHV Avo.va Hmm.omv Awo.mHv Hmv.mHV Amm.mHo oHuuo EHJCCH.SHmw mH.Hm o>.mm mm.Hm mm.vm mm.vm 0®.om om.mw vv.om mb.mh ocsnmxonsmssfié 0&0 Hmv.0mv va.omv Hmm.mmo HvH.mHv Amm.mHv HHN.NNV Hom.Hmv Hmm.omv AMH.mHV mosh: um©H0i mflxom N0.0m mm.mm mv.mw vm.ob oo.mm mv.bm mm.mm mb.mm mm.mm -c?Hmcd%:3 chtzuo mchcH Hm0.HNV Amo.mHo Amm.mHv Hmm.th ANH.vHV Amv.mHv Hbm.mHV Hmh.mHo Hmm.bHo ummH0: mfion.o mv.mw mm.v> ov.Hb Hb.mh wb.bb ov.m> mo.mh hm.mh wm.mw -::H.%::cssH: m>fix0 .HOu huHHHno “30% CH 30> who cHouHmo 30m. coHummSO m m b w m 0 m N H m>mz .A©.ucouv mH mHDMB APPENDIX J APPENDIX J Data Coding Sheet Team I.D. number Previous contact percentage (cumulative) Previous contact percentage (most recent) Previous hitting performance (subjective) Game 1 self-efficacy Game 1 subjective performance Game 1 contact percentage Game 1 contact percentage (cumulative) Game 1 effort Game 1 anxiety (overall) Game 1 cognitive anxiety Game 1 somatic anxiety Game 2 self-efficacy Game 2 subjective performance Game 2 contact percentage Game 2 contact percentage (cumulative) Game 2 effort Line 1 116 6-9 10-13 14-15 16-20 21-23 24-27 28-31 32-33 34-35 36-37 38-39 40-44 45-47 48-51 52—55 56-57 UJH ‘ ‘ ubN college high school mm Game 2 anxiety (overall) Game 2 cognitive anxiety Game 2 somatic anxiety Game 3 self-efficacy Game 3 subjective performance Game 3 contact percentage Game 3 contact percentage (cumulative) Game 3 effort Game 3 effort Game 3 anxiety (overall) Game 3 cognitive anxiety Game 3 somatic anxiety Game 4 self-efficacy Game 4 subjective performance Game 4 contact percentage Game 4 contact percentage (cumulative) Game 4 effort Game 4 anxiety (overall) Game 4 cognitive anxiety Game 4 somatic anxiety Game 5 self-efficacy Game 5 subjective performance 117 58-59 60-61 62-63 64-68 69-71 72-75 76-79 80 2-3 4-5 6-7 8-12 13-15 16-19 20-23 24-25 26-27 28-29 30-31 32-36 37-39 0 Game 5 contact percentage Game 5 contact percentage (cumulative) Game 5 effort Game anxiety (overall) somatic anxiety 5 Game 5 cognitive anxiety Game 5 6 Game self-efficacy Game 6 subjective performance Game 6 contact percentage Game 6 contact percentage (cumulative) Game 6 effort Game 6 anxiety (overall) Game 6 cognitive anxiety Game 6 somatic anxiety Game 7 self-efficacy Game 7 self-efficacy Game 7 self-efficacy Game 7 self-efficacy Game 7 self-efficacy Game 7 subjective performance Game 7 contact percentage Game 7 contact percentage (cumulative) 118 40-43 44-47 48-49 50-51 52-53 54-55 56-60 61-63 64-67 68-71 72-73 74-75 76-77 78-79 80 5-7 8-11 12-15 .,11 Game 7 effort Game 7 anxiety (overall) Game 7 cognitive anxiety Game 7 somatic anxiety Game 8 self-efficacy Game 8 subjective performance Game 8 contact percentage Game 8 contact percentage (cumulative) Game 8 effort Game 8 anxiety (overall) Game 8 cognitive anxiety Game 8 somatic anxiety Game 9 self-efficacy Game 9 subjective performance Game 9 contact percentage Game 9 contact percentage (cumulative) Game 9 effort Game 9 anxiety (overall) Game 9 cognitive anxiety Game 9 somatic anxiety 16-17 18-19 20-21 22-23 24-28 29-31 32-35 36-39 40-41 42-43 44-45 46-47 48-52 53-55 56-59 60-63 64-65 66-67 68-69 70-71 APPENDIX K APPENDIX K Raw Data 4527407930500050766702010000800 321913 0861104006670778102413110927805010000806092712150938905006670795102512130 92780200333076210291712 49813100010000305056 0577808006670750 2512130611104005000625 2615110611105010000667 2010100594404007500692 2212100644403005000667 2313100 577803000000588 2414100555607010000650 180909061110801000069610191009 456960750100008081110601000078307261313 07389 066670200667075905452025057220801000078l 201010 051110601000077806200911 42374097010000307444040100009720929l61307389010100009750835201508278050100009770 9271809 0788903010000982093322110716703010000983 3422120 62220300000096610402317072220201000096810372512066670801000097010301218 46082073910000305556020066707310845242104556020100007410945242105667030100007670 94926230666702006670758094926230705605010000795 4523220672201006670786086633330 4611050l0000795095830280777807007500792093920l9 446840943100008 06278080100009460 8291712062220601000095008331914067780301000095607382216 0666704010000958063620160705605010000961073219l3 42873l000100003066670601000100008251609057780401000100004231409 061110501000100004211209 061670500667097803210912 46958076910000506111020033307380633201305556050075007390731181306611010100007550 6342311 0611101000000740 4027130 56110300250070407422616066670200333068405432716 417220667050004 04889030050006000540241605278070000005000 7301812 051110800500050008301317 14303909130667040755604010000920074122190916701007500897085421330872207010000909l 03820180927808010000914102714130944406010000919102613130955610010000923101910090 94441001000092910180909095560500500090908221012085560901000091510180909 418300632100004061l1050100006820828181006611060066706800729181107222060100007140 72616100705607006670710082213090683309008330730082515100680007010000750082717100 68890700667074408251510 075000700750074507261511 401430844100008083330600500082409251213075560800500080610230914083330701000081ll 01909l00822209010000829102409150877805005000800104018220855607010000813102410140 70560800667080410291514076670901000081510221012074440700500079309321913 46475033310000306444 05444060000003330945232205333 06722 0 655604005000364 27151206833 4462509091000060750006010000917093019]1083330801000092510281711 07889080100009321024150908278040080009181024l31l 075000501000092310251510073890301000092910261610 3708908l8l00004062220800500076908351916070000501000081307271809071l1040100008420 72414100766708010000870092314090761103007500852072715120777808010000862092314090 7667090100008670928151307722 073330901000087509241509 39494l000100006076670801000100010352l14 1220 1121 0727805010001000 382018 0 7778080100009581027161108111060100009631040231707778 31839075000000505444 05500 05889020060006670 454272704667040050006360551272404778 0450006007500667 5027230 4556070l000068807432320047220200500066703482523045560500500065005402020 39961091710000506611040075008750832181407667 0777804006670842 3016140872207010000870 3117140 91110100500084007331716088330800667082110251l14092220501000082809231013 3513610001000050794408010001000 22121007444060100010000118090907889050100010000 8180909073330501000100006210912 0 72220601000100006210912072220701000100006180909074440601000100006180909 357000583l0000606000030025005000748252305778040033304740649321705ll110006000500 5226260688909005000500104627190738908005000500093724130750007010000559073322110 079440300500055306592930 39134000000000205611 0805607005000167093016140861109010000444102109120 8556090100005001022091308889070100006150725l31209167060100006670720091l 3634008461000040383308007500824 06944090100008640 9241410073330901000088009371819 35922083310000406944050100008671033221108000020100008950849252406333090100009171 02817110811104010000926093419150744405006670900103319140711103005000875093823150 68890701000088610412219070000100000081608483018016670101000082909503218 329111000100006 081l1030100010000723111208444050100010000 735171808444 0933305007500960083617190911102510000964063718190 9000030100009660337181908611050100009700835181708722 367880923l0000707222050075009230728171108667070100009000728171108056060075008750 822121008000070100008890823l4090727803003330833054022180655608010000839102915140 78330100750082904351520063330801000083808221309086110200333080002311813 355151000100008060560601000100008362016 057780601000100008382117 0 50560401000100007401921055560400000096407412318047780901000096908321814 34032 08 06944020100010000836201606056080100010000 9211209066111001000100009231l12 O 62780650667092309211011 38205080007500206722 0738905010000857082314090761l 0783303010000889102112090750002005000850102614120738910010000870101809090 8611045075008520926l4120822207005000828072614120861l1000667081310180909 214840831066704 08500090050008150926151108222080100008300 92112090861107008000828082112090800007510000837082112090827808010000846081910090 84440400800084407251312081l1080100008500926161007944030066708450729181l 24561074405000607944 0811110010000781 29161308444070100008001 0221309081111001000081110291613081l106510000829102112090816710010000833102513120 8111 08278 2794508321000080900007010000839083928l109667040100008480638211709667045066708440 93016140955609010000848102516090966706010000858102314090961107010000864092816120 92780550600085510402614095560600667085209352015092220701000085410362214 24688080010000508667065083308031030191108500040100008090827171007778060100008190 8 0816703010000848 3120110877809010000827093120110833304505000818073421130 7778050040007 93 10 0844407005000786103421 13086110301000079109 23661089110000808333045088308881021120908222025100008930821120908667045100008991 01910090838902006000888082112090850004010000894102112090800006508750893101910090 1222 78890200833089110221210081l10501000089310191009084440100750089010211110 ' 231870904l0000708389065066708920936231308667085087508911031181309056085100008961 02515100900003010000900104120210866709010000905102715120922208510000908102414100 92780100857090610372215092220601000090710261511093890701000090910251510 26210094107500606944 07056040100009500930191107444060050008750 935201507556 0688901010000889093720170783307010000897092313100 7500 07667 08000 25715090310000607667040060008890833221108167040083308860932211107667060100008931 03221110783308010000897103018120788906010000902083219130822204010000906093321120 761104010000909093119120783303510000910103019]1080560200500090510321814 221250821100004081 1 1 083890301000083309 08056080100008440 9 08111 0816709010000854102616100877807007500846092615110 9056 09111070100008520928141408944 27840081310000707444050050007901027111609111050075007881019100908722010100008000 92616100861101010000806082515100894403010000814082210120955607007500812101809090 9667055080008111020101009278 0950001005000811 191009 24597081410000407722050066708060936201608667010033307800836152107944070100007911 03018120822208510000800092716]10866708010000813102917120844404508000812082513120 766705008000812093417170816704010000814 2814140833305010000818 251213 29444085406670605833065100008680947281906889 07722040100008750 9352213063890901000087710361917070560501000088109231409 06167 07667 2706008380750080877803505710821062913160883305007500819 1909100961104010000825 2111100961106007500823 2412120966704006000814 3318150955608010000818102513120 98890950800081810241113099440700750081609191009097780501000081908211110 12668095307500707389080100009561030l81207667050100009581037211607611100090009541 030191108333 079440601000095610241509081l101510000959102415090 9056 0983306008330954102213090966702010000955 211209 11620069l03330407056050050006840743232008333045100006960830181208778020040006820 93917220877806010000691092l11100766701003330681023221110916702008890697102415090 7889060033306881035211406944050066706881039211809000070050006851026151l 15363083910000608444035080008371022111109222060100008461023111209222050077808421 02111100950001010000846102712150944404510000853102311120883307009090861101809090 750001010000861102410140877801010000866102310130838901010000868 180909 10550088605000508222040100008910719100908222050085708900823l11209667050100008970 6190910096670501000089905251411086110651000090308180909085560800833090110180909 0950002510000903063013170777801006670899073l1615 17759067808000708167055066706671021101108111030033306670823111207667040050006541 0231112082780100400064307241410079440100333062708401921076670300667062810371720 07667 18500064506000404000 05056 05667060100006760 93221120488904010000732103625l1 0 57220200750073308442519 04444 14709082105000607444090083308221034211306389080100008311036221406000100100008401 03319140644410010000843 3819190627807508570843104226160616706008000836094124170 544406007500833103623130661104010000840104124170650005010000842 442618 15781089310000607500020100008971028161208389060100009001027161107611030085708980 92213090800006010000901102111100800003010000904081910090805604508890903102112090 7667 04000889 2412120722204508570888082514110750005006670884 231310 15466086110000706889035100008680738241406167085100008720930181206944030075008660 73421130783305508330865092818100738901505000848095028220688904509000851103322110 755606010000854082111100700006510000859102917120766709010000860 291811 1323 11993083903330409611055100008471032181409500070060008371026151109778035100008470 83018121000006508000846093l171409444080075008430925141l1000007508890846102212100 98890801000085110231310098330601000085508251510097220301000085608401822 10858082305000206222 07778070066708150835181707611080100008240 922111108222035080008220726141208556020066708160721ll100833305008330827082010100 85000801000083509190910081670500667082409210912078330301000082807201010 15853077810000306889 07278 08722 08778060100007851029191007500 0833301006670779103321120 8278020050007711033221107444 08000 L IST OF REFERENCES LIST 01“ REFERENCES Bandura. A. (1969). Erincip1ss_Qf_benayior_modification. New York: Holt, Reinhart, & Winston. Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological_8eyiew 84. 191-215. Bandura, A. (1978). Reflections on self-efficacy. In S. Rachman (8d.), Adyances_in_behayiour_research_and_ therapy (Vol. 1. pp. 237-269). Oxford: Perganon. Bandura. A. (1986).SQcia1_foundations_of_thonght_and_ asL1on1__A_social_cpgnitiye_theory Englewood Cliffs. NJ: Prentice- -Hall. Bandura, A. (1990). Perceived self— -efficacy in the exercise of personal agency flournal_of_Applied_Sport_ Psychology .2. 128- 163. Bandura, A., & Adams, N. E. (1977). Analysis of self- efficacy theory of behavioral change. Cognitiyg_1herapy and_Bessarch. 1. 287- 308. Bandura, A., Adams, N. E. Hardy, A. B. & Howells, G. N. (1980). Tests of the generality of self- -efficacy theory Cognitiye_Therapy_and_Eesearch .4. 39- 66 Bandura, A., Reese, L., & Adams. N.E. (1982). Microanalysis of action and fear arousal as a function of differential levels of perceived self— —efficacy. Journal_gf 2srsonality_and_Social_Esycholpgy. 43. 5- 21 Barling, J.. & Abel, M. (1983). Self— -efficacy beliefs and performance Cpgnitiys_Iherapy_and_Besearch .1. 265- 272. Baron, R. (1970). Attraction toward the model and model' 5 competence as determinants of adult imitative behavior. Jonrnal_of_2ersQuality_and_Social_Esychelegy. 140 345- 351. Baron, R.M., & Denney, D.A. (1986). The moderator—mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Qonrnal_of_Eersanality_and_Social_Esychology. 51. 1173— 1182. Borkovec, T. D. (1976). Physiological and cognitive processes in the regulation of fear. In G.E. Schwartz and D Shapiro (Eds. )- Consciousness_and_self; regulationi__Adzances_in_ressarch. New York: Plenum. 124 125 Brown. I., & Inouye. D.K. (1978). Learned helplessness through modeling: The role of perceived similarity in competence. ' ' Esmhology. .16, 900-908. Carroll, w. R. & Bandura. A. (1985). Role of timing of visual monitoring and motor rehearsal in observational learning of action patterns W. 11, 269- 281. Courneya. K. S. & Chelladurai, P. (1991). A model of performance measures in baseball. Journal_gfi_fipg;;_and W 11. 16- 25. Dweck, C. S. (1978). Achievement. In M. E. Lamb (Ed. ), WW (pp 114- 130) New York: Holt. Rinehart & Winston. Feltz, D. L. (1982). The effect of age and number of demonstrations on modeling of form and performance. W. 5.3.. 291- 296. Feltz, D.L. (1984). Self-efficacy as a cognitive mediator of athletic performance. In W.F. Straub (Ed.), W's-1131191991 (pp. 191-198). Lansing. NY: Sport Science Associates. Feltz, D. L. (1988a). Self— confidence and sports performance. In K. B. Pandolf (Ed. ). W .science_reyiews (Vol. 16, pp. 423- -457). New York: Macmillan. Feltz, D. L. (1988b). Gender differences in the causal elements of self- -efficacy on a high avoidance motor task. W 1.0. 151- 166. Feltz, D. L. (1992). Understanding motivation in sport: A self-efficacy perspective. In G. C. Roberts (Ed.). ' ' ' Champaign, IL: Human Kinetics. Feltz, D. L. , Bandura, A. & Lirgg. C. D. (1989. August). Perceived collective efficacy in hockey. In D. Kendzierski (Chair) . WW Symposium conducted at the meeting of the American Psychological Association, New Orleans, LA. Feltz, D.L.. Landers. D.M.. & Raeder, V. (1979). Enhancing self-efficacy in high avoidance motor tasks: A comparison of modeling techniques. Journal_gfi_$pgzg ESlQthQQX. 1. 112-122. 126 Feltz, D.L., & Mugno, D. (1983). A replication of the path analysis of the causal elements in Bandura's theory of self-efficacy and the influence of autonomic perception. WW. 5. 263-277. Feltz, D.L., & Riessinger. C.A. (1990). Effects of in vivo emotive imagery and performance feedback on self— efficacy and muscular endurance. Journal_gfi_fipgrt__§nd WNW-1.2.. 132- 143. Fitzsimmons, P.A., Landers, D.M., Thomas, J.R., & van der Mers. H. (1991). Does self-efficacy predict performance in experienced weightlifters? Research W. 62.. 424-431. Gayton, W.F., Matthews, G.R., & Burchstead. G.N. (1986). An investigation of the validity of the physical self- efficacy scale in predicting marathon performance. WW3- 61. 752- 754. George. T.R. (1988). Wk o- ‘0 u... - «new. '0 ‘ 0 ‘0 -. ‘ .0- ‘9- .' ' ° -' exploratory_study. Unpublished master’ 3 thesis, Miami University. Oxford, OH. George, T.R., Feltz, D.L., & Chase, M.A. (1992). Effects of model similarity on self-efficacy and muscular endurance: A second look. _1Qnrnal_gfi_spg;;_and W. 1.4. 237-248. Gill. D- L. (1986). 2W Champaign, IL: Human Kinetics. Glass, G.V. & Hopkins, K.D. (1984). Stanisgjga] m§§thS in W. Englewood Cliffs, NJ: Prentice-Hall. Gmelch, G. (1972). Magic in professional baseball. In G. Stone (Ed.). WW: (pp. 128-137). New Brunswick. NJ: Transaction. Gould. D., & Weiss. M. (1981). The effects of model similarity and model talk on self-efficacy and muscular endurance. MW. 3.. 17-29. Gould. D., Weiss. M.R.. & Weinberg. R. (1981). Psychological characteristics of successful and nonsuccessful Big Ten wrestlers. Qournal_o£_&port Esychology. 1. 69-81. 127 Griffin. N.S.. & Keogh, J.F. (1982). A model for movement confidence. In J.A.S. Kelso and J. Clark (Eds.), The ' ' (pp. 213-236). New York: Wiley. Haney. c.J. (1991). W W sport_gompetitions. Unpublished doctoral dissertation, University of British Columbia. Vancouver. Harter, S. (1978). Effectance motivation reconsidered: Toward a developmental model. Human_peyelopmen;, 21, 34—64. Highlen, P.S., & Bennett, B.B. (1979). Psychological characteristics of successful and nonsuccessful elite wrestlers: An exploratory analysis. Journal_gfi_$pgrt Esyshology. 1. 123-137. Highlen, P.S., & Bennett, B.B. (1983). Elite divers and wrestlers: A comparison between open- and closed-skill athletes. 188W. 5. 390-409. Kavanagh, D.J., & Bower. G. H. (1985). Mood and self- efficacy: Impact of joy and sadness on perceived capabilities WW 2, 507- 525. Kavanagh, D.J., & Hausfeld, S. (1986). Physical performance and self-efficacy under happy and sad moods. Journal_gfi W. 8. 112-123. Lan, L., & Gill, D.L. (1984). The relationships among self- efficacy, stress responses. and a cognitive feedback manipulation. MW. 6. 227-238. Lee, C. (1982). Self-efficacy as a predictor of performance in competitive gymnastic-s. W. A. 405-409. Lirgg, C.D. (1991). Gender differences in self— confidence in physical activity: A meta- analysis of recent studies. W 1.1. 294-310. Lirgg. C.D., & Feltz. D. L. (1991). Teacher versus peer models revisited: Effects on motor performance and self- -efficacy 8W Sport. 6.2.. 217- 224 Maccoby. E.E., & Jacklin, C.N. (1974). Ihe_psyghglggy_gfi §gx_dififierengea. Stanford. CA: Stanford University Press . 128 Mahoney, M.J., & Avener. M. (1977). Psychology of the elite athlete: An exploratory study. Cognitiye_1herapy_and Research. 1. 135-141. Martens, R., Burton. D., Vealey, R.S.. Bump, L.A.. & Smith, D.E. (1983). W Unpublished manuscript, University of Illinois at Urbana-Champaign. Martens, R., Burwitz, L., & Zuckerman, J. (1976). Modeling effects on motor performance. Researgh_QuazterlyJ 41, 277-291. Meyers, A.w., Cooke, C.J.. Cullen. J.. & Liles. L. (1979). Psychological aspects of athletic competitors: A replication across sports. Cognitiye_1herapy_and 8.88881811- .1. 361-366. McAuley, E. (1985). Modeling and self-efficacy: A test of Bandura's model. W. l. 283- 295. MCAuley. E. (1990. June). WWI-i; . .01 ' ‘0. H -.- 7‘ 0‘0. '0 '9 .909 .0. Paper presented at the meeting of the American Psychological Society, Dallas. McAuley, E.. & Gill. D. (1983). Reliability and validity of the physical self-efficacy scale in a competitive sport setting. Whom. 5. 410-418. McCullagh. P. (1986). Model status as a determinant of observational learning and performance. Journal_gf W. 8. 319-331. McCullagh. P. (1987). Model similarity effects on motor performance. MW. 8. 249-260. Miller. R.G. (1966). 81W New York: McGraw-Hill. Ness: R G-: & Patton. R.w. (1979). The effects of beliefs on maximum weightlifting performance. gggnjgjxe Iberanx W11. .1. 205-211. Nicholls, J.G. (1978). The development of the concepts of effort and ability, perceptions of own attainment, and the understanding that difficult tasks require more ability. W. 42. 800-814. Nicholls...J.G. (1980). WWW Paper presented at the University of Bielefeld, West Germany. 129 Pedhazur. E.J. (1982). HEW W. New York: Holt. Rinehart & Winston. Schunk, D. H. (1981). Modeling and attributional effects on children’ s achievement: A self— -efficacy analysis. W 13. 93-105. Schunk, D. H. (1984). Self— efficacy perspective on achievement behavior W.1&. 48- 58. Schunk, D. H. & Cox. P. D. (1986). Strategy training and attributional feedback with learning disabled students. MW 18. 201- 209. Schunk, D. H., & Gunn. T. P. (1986). Self— -efficacy and skill development: Influence of task strategies and attributions. W. 19.. 238- 244. Shaw, J.M., Dzewaltowski, D.A., & McElroy, M. (1992). Self- efficacy and causal attributions as mediators of perceptions of psychological momentum. Jon;nal_gfi_fipgz§ W 1.4. 119- 133 Shelton, T. O. & Mahoney, M.J. (1978). The content and effect of 'psyching- up" strategies in weightlifters. WW. 2.. 275—284. Taylor, C.B., Bandura. A., Ewart, C.K., Miller, N.H., & DeBusk, R. F. (1985). Raising spouse' s and patient' 3 perception of his cardiac capabilities after clinically uncomplicated acute myocardial infarction. American W- Vealey, R. S. (1986). Conceptualization of sport-confidence and competitive orientation: Preliminary investigation and instrument development. Journal_gfi_fipgzg W. 8. 221-246. Weinberg, R.S.. Gould, D., and Jackson, A. (1979). Expectations and performance: An empirical test of Bandura's self-efficacy theory. lgn;nal_gfi_§pgz§ 88mm. 1. 320-331. Weinberg, R.S.. Gould. D., and Jackson. A. (1980). Cognition and motor performance: Effects of psyching- up strategies on three motor tasks. Cognitiyg_1hgrapy W. 4. 239-245. 130 Wilkes, R.L.. & Summers, J.J. (1984). Cognitions, mediating variables, and strength performance. Jon:nal_gj_fipgrt 28181181981. 8. 351-359. STATE UNIV. L “MI"HHUH”WWI 891" 9300 ”11111)