X55. ‘ 5 . u‘l‘ F W "X‘v.~hlf»\ -¢Wl~‘-)~ ‘rvw- 1W 5' I'tr - ,u l “.2 . ' .5 .1 ‘1 5., . . I ‘.‘ ’5 | ‘ n ‘ a 1! ::5 . $Y5.5 .5. “X 1' "‘15. I X . I' .. , I | 5 F": '."X .51, I1 :‘X‘Wc ’5 ! ‘.. ' a. 1 '1 - 5 -" 5' "'IX 5 ‘ll '5‘ " '< "I' 5‘ 5 1‘ 5 5 l .u'1 5 5' ' ‘ I '1' . I \5 5 '5'." 5 1 X'I' "' .5 ‘ I , X l' 55-. " X X .1 " m',' 5; 1 . 1 ‘5. 5. _ ‘ ""55 ' ]' '1'1 " 1 -\ 5. 'X XX ’X ‘XI I . , '1 " 5:13.521. ‘. I5 51 ',' 1:. 1-1, 55515119 1"1 .1. 5 5&5}. 1‘" 5.1.5555 .3‘5. __ "if . 1:15“ I I55“ 55-; \I‘J .4 5 ' a“ 5155:1251 .‘ 5.5” ‘5 I In" 5' .0“ ”:5. 1 l H. 11”” 15 5 5| . '; ‘_ 5' '- | 511.; 51:56"? {135 '5,:)\Rf"\ 51W] 55.". ‘5'... 3 . ' , H". .. .. " X X , . . .1 X. “I. .g ‘ <0“ 5" 5‘ . 1 l‘ 3" 5:" ' I 7.55 ' I ' 5 5.55“” ' I 15-:1 £51... 4 ( '|’I ... ll . -. I XX. - 1:51 1 1 ‘3955'55'0 JX' ' ' .35 ' 5 I .1, X5“? 5 "MM ‘- 1'X \' ' X‘ " ‘ I"'X ' X'l " I' " r a n' '5' ' 5 I" ' ”1.. '5' ‘5 151' .. XX ‘ X l I l5 . 5 5“ . .1 I X 5 ' f ”'55 ‘5' I '.'.' ' '5?" "' _"' mm]; 515' 5 " '15-1'1"5X 35:53.». 55‘ . '. 5'55 .1551" 5 '15. ' X - 5" ‘ ulX 9.. Jun “5' .5 ll ..XI I u . 1. X .XX' ‘ 5. "5‘... ."|- 'L .. 55\l 1. , ’ , 1 ,' ‘ ., 1 1., , I ' g . flu -‘ 1 .. ." '~ ‘ X>‘5'5"5|' ‘ .1 .- ="‘-.1s.:‘?~“'-:; ’ 5 " ' " :.. 1.5015 5"“1.£ '51., X X ‘ 1-: 'I X ' l‘fiéfik 15‘ .. .,v‘ ‘7 N r‘ .1 . 5."1""_-1X\" ' :111‘595'. 5555' a. «H r_... :_ _‘ __'. _ - f:_ - -_ r 5435: 0". .N.‘\ 1 1 )5 , u ' I’ II 3.5.1"; "' ".55. ”‘11 ’1‘ .‘k‘: ‘ ‘: . ._‘.A-.U' .X 1.5155” “yd ._' u‘o'Jl 518355". I' . 5 "ih‘kauéffir '9 .5 “7‘ .“ ,5. .h 153.15} 55 5/15" l1. '51.. 111': XX': 15:55" " 1 év ’til‘fl w "6:1?" 5,5 " :-'1. . . 5,155 “:5”? «13' 55 " '95"; @5515 ' ‘ 555553 ' .75 ‘53 ding-5:" at}: 1. 555’“ 15151.55. 5’5 "h ' ' 2- : "'5” - .. .,. ’3 Pi « (115511-21. '5'" . g3 - 0 '- .. _ ' "'15." i} 55"}? 9%“? ’5‘“; ' '1. u n': 5. ‘4 ‘ ' " '1‘"'X.'1:¥5.'{"'51§-'35 “52%?! t .5” '1 I. TH ESlS LIBRARY M 1cm gm: State University ——— v This is to certify that the dissertation entitled DEVELOPMENT AND EVALUATION OF A SELF-TALK ASSESSMENT INSTRUMENT FOR TENNIS PLAYERS presented by Nancy Lynn Van Noord has been accepted towards fulfillment of the requirements for Ph oD 0 degree in Health and Phys 1C3]. Education _ ,2 -3 ( 3 r | a' (f;k“r_’_, ‘4' n‘ “4 g 41' ~ ”5’3 DEVELOPMENT AND EVALUATION OF A SELF—TALK ASSESSMENT INSTRUMENT FOR TENNIS PLAYERS By Nancy Lynn Van Noord A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Health and Physical Education 1984 3,; 3 3 you: copyright Nancy Lynn Van Noord 1984 DEVELOPMENT AND EVALUATION OF A SELF-TALK ASSESSMENT INSTRUMENT FOR TENNIS PLAYERS By Nancy Lynn Van Noord AN ABSTRACT OF A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Health and Physical Education 1984 ABSTRACT DEVELOPMENT AND EVALUATION OF A SELF-TALK ASSESSMENT INSTRUMENT FOR TENNIS PLAYERS By Nancy Lynn Van Noord The purpose of this investigation was to design a self-talk questionnaire (STQ) which could validly and reliably assess the type and frequency of a tennis players' competition-related self-talk as it related to competitive psychological effectiveness. A valid and reli— able Sl-item self-talk questionnaire (STQ-3) was developed which evolved from the sequential construction and subsequent administra- tion of the STQ-l and STQ-Z. A total of 302 male and female tennis players, ranging in skill from intermediate to world class caliber, from Florida, Michigan, and Texas, completed either the STQ-l, the STQ—Z, or the STQ-3. Based on the results of a factor analysis, the investigator discovered that the type of self-talk emitted by these players in the competitive setting, could be classified as either situation Appropriate or Inappr0priate. Furthermore, the type and frequency of self-talk emitted by psychologically effective players was significantly differ- ent from that voiced by psychologically ineffective players. Although both groups of players engaged in a moderately high frequency of Nancy Lynn Van Noord Appropriate self-talk, the ineffective players voiced a significantly higher frequency of Inappropriate self-statements. Implications of these findings and suggestions for future research are also discussed. ACKNOWLEDGMENTS I would like to acknowledge the efforts of a number of people who provided me with their expertise, encouragement, and support during the development and completion of this dissertation. Administrators and physical education faculty of Calvin College who made every effort to give me sufficient time and support to enable me to complete my doctoral degree. Dr. Annelies Knappers, my advisor, mentor, and long-time friend, who has nurtured in me feelings of self-confidence and self-respect, who has challenged me since I was an undergraduate to develop my God-given talents, and who helped me to grow as a researcher and scholar. Dr. Dan Gould, Dr. Bill Hinds, Dr. Judy Tant, and Dr. Dozier Thornton, my doctoral committee members, who gave me many beneficial suggestions which improved the quality of my dissertation. Jan Vredevoogd, my statistical consultant, whose guidance was invaluable throughout the course of my dissertation work. Nancy Heath, my typist, who volunteered to do everything in her power to help me meet my deadlines, and who gave me a bountiful supply of Christian support. Andy and Gladys Van Noord, my parents, who assisted me finan- cially, and who willingly put their own work aside to help me with coding and proofreading tasks. ii Joan Nelson, my karate instructor, role model and friend, who through her enthusiasm and professional example, helped me to become a better teacher and scholar. Dianne Ulibarri, Judy Walton, and my fellow karate-kas, who through their loyal companionship helped me to keep a proper perspec- tive on my dissertation work, and who nurtured and sustained in me the fortitude to earn my doctoral degree. iii TABLE OF CONTENTS Page LIST OF TABLES . . . . . . . . . . . . . . . . V11 LIST OF FIGURES . . . . . . . . . . . . . . . . ix LIST OF APPENDICES . . . . . . . . . . . . . . . x Chapter I. INTRODUCTION . . . . . . . . . . . . . . . 1 Statement of the Problem . . . . . . . . . . 7 Significance of the Study . . . . . . . . . . 7 Assumptions Underlying the Study . . . . . . . 9 Scope of the Study . . . . . . . . . . . . 10 Definition of Terms . . . . . . . . . . . . 10 II. REVIEW OF RELATED LITERATURE AND RESEARCH . . . . . 13 Influence of Self-Talk on Behavior . . . . . . . 13 Age of Subjects . . . . . . . . . . . . 14 Type of Experimental Setting . . . . . . . . 15 Type of Intervention Program . . . . . . . . 16 Dimensions of Self-Talk . . . . . . . . . . 20 COflClUSionS e o o o o o o o o o o o o o 25 Assessment of Self-Talk . . . . . . . . . . 26 Recording . . . . . . . . . . . . . . 29 Sampling . . . . . . . . . . . . . . . 33 Production . . . . . . . . . . . . . . 36 Endorsement . . . . . . . . . . . . . . 37 Conclusions . . . . . . . . . . . . . . 40 III. PROCEDURES . . . . . . . . . . . . . . . 44 Phase I: First Stage in the Development of STQ’B (STQ-l) o o o o o o o o o o o o o 44 Construction of STQ-l . . . . . . . . . . 45 Selection and Construction of Player Rating Scales . . . . . . . . . . . . . . . 46 Selection of Sample . . . . . . . . . . . 47 iv Chapter Page Administration of STQ-l . . . . . . . . . 50 Collection of Player Skill (NTRP) and Competitive Psychological Effectiveness (CPE) Ratings . . . 51 Compilation of Self—Statements . . . . . . . 51 Phase II: Second Stage in the Development of the STQ-3 (STQ-Z) . . . . . . . . . . . . . 54 Elimination of Redundant and Inappropriate Responses . . . . . . . . . . . . . 55 Classification of Statements . . . . . . . . 56 Employment of Raters to Assess Remaining Statements . . . . . . . . . . . . . 59 Construction STQ-2 . . . . . . . . . . . 60 Revision of CPE Rating Scale . . . . . . . 62 Pilot Testing of STQ-2 . . . . . . . . . . 62 Selection of Sample . . . . . . . . . . . 62 Administration of STQ-Z . . . . . . . . . 63 Collection of Players Skill and Competitive Psy- chological Effectiveness Ratings . . . . . . 64 Phase III: Third Stage in the Development of the STQ-3 . . . . . . . . . . . . . . . 68 Examination of Order Effect . . . . . . . . 68 Analysis of STQ-Z Responses . . . . . . . . 69 Construction of STQ—3 . . . . . . . . . . 91 Selection of Sample . . . . . . . . . . . 92 Administration of the STQ-3 . . . . . . . . 94 Collection of Player NTRP and CPE Ratings . . . 95 IV. FINDINGS AND DISCUSSION . . . . . . . . . . . 96 Demographics of the STQ-3 Sample . . . . . . . 96 CPE and NTRP Ratings of the STQ-3 Sample . . . . 96 Psychometric Properties of the STQ-3 . . . . . . 101 Construct Validity . . . . . . . . . . . 101 Reliability . . . . . . . . . . . . . 106 Analysis of Results . . . . . . . . . . . 107 Summary of Results . . . . . . . . . . . . 113 Discussion and Implications . . . . . . . . . 114 V. SUMMARY, CRITIQUE, AND FUTURE DIRECTIONS . . . . . 120 Summary . . . . . . . . . . . . . . . 120 Critique of STQ-3 and Future Directions . . . . . 127 Instrument Variation . . . . . . . . . . 127 CPE Design Difficulties . . . . . . . . . 130 Corollary Research . . . . . . . . . . . 132 Chapter Page APPENDICES O O O O O O O O O O O O O O O O O 135 LISTOFREFERENCES . . . . . . . . . . . . . . 190 vi 10. 11. 12. 13. 14. 15. LIST OF TABLES Page Phase I Frequency Distribution of Ages of Players and Years of Tennis Playing Experience . . . . . . . 52 Frequency Distribution of STQ-l CPE Ratings . . . . 53 Frequency Distribution of STQ—l NTRP Ratings . . . . 53 Sequence of STQ-l Item Elimination for Construction of STQ-z O O O O O O I O O O O O I O O O O 55 Distribution of STQ-l Self—Statement Classifications . 58 Phase II Frequency Distribution of Ages of Players and Years of Tennis Playing Experience . . . . . . 65 Frequency Distribution of STQ-2 Coach Assigned CPE Ratings 0 O O O C O O O O O O O O O O O 67 Frequency Distribution of STQ—Z Coach Assigned NTRP Ratings . . . . . . . . . . . . . . . . 68 Results of Wilks' Lambda to Assess the Effect of Cate— gory Order on Player Responses . . . . . . . . 69 Sequence of STQ-Z Statistical Analyses and Item Elim— ination for Construction of STQ-3 . . . . . . . 71 STQ-Z Statements which Significantly Discriminated Between the CPE Criterion Groups . . . . . . . . 74 STQ-Z Statements which Significantly Correlated with the Five Levels of CPE . . . . . . . . . . . 77 Omitted STQ-2 Items Based on Nonsignificant Results of Discriminant and Correlational Analyses . . . . . 79 Factor Loadings of Remaining 58 STQ-Z Items on Factors I and II C O O O O O O O O O O O O O O O 82 Cronbach's Alpha Coefficients for the Remaining 58 STQ-Z Items . . . . . . . . . . . . . . . 87 vii' TABLE 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. Page Cronbach's Alpha Coefficients for STQ-2 Items Selected on the Basis of Discriminant and Spearman Rank Order AnalySis O O O O O O O O O O I O I O O 0 88 Cronbach's Alpha Coefficients for Remaining 52 STQ-Z Statements 0 C O O O O O O O I O O O O O 91 Distribution of STQ-3 Inappropriate and Appropriate Items 0 O O O O O O O O O O O O O O O O 93 STQ-3 Frequencies and Percentages for Gender and Player Type O O O O O O O O O O O O O O O O O 97 Descriptive Statistics on STQ-3 Personal Player Data . 97 Comparison of Descriptive Statistics on Personal Player Data for the STQ-l, STQ-Z, and STQ-3 . . . . . . . 98 STQ-Z and STQ-3 CPE Score Distribution . . . . . . 199 STQ-Z and STQ-3 NTRP Score Distributions . . . . . 100 A Comparison of STQ-Z and STQ-3 Percent Variance and Eigen Values for Factors I and II . . . . . . . . 105 Cronbach's Alpha Coefficients for Final 51 Items Based on STQ-Z and STQ-3 Data . . . . . . . . . . . 106 Means and Standard Deviations for Player Responses to Inappropriate and Appropriate Factor Items . . . . . 108 Discriminant Function Results for Inappropriate and Appropriate Factor Items . . . . . . . . . . . 112 Results of Discriminant Function Analysis on 118 STQ-Z Items 0 o o o o o o o o o o o o o o c o 178 Results of Discriminant Function Analysis on 50 STQ-2 Items 0 o o o o o o o o o o o o o o o o 180 viii LIST OF FIGURES Schematic Representation of No Basic Self-talk Dimensions . . . . . . . Response Pattern to 32 Discriminating Inappropriate STQ-3 Items by Psychologically Ineffective and Effective Players . STQ-3 95% Confidence Intervals for CPE Group Responses . . . . . . . Proposed Revision of CPE Scale ix Page 20 103 113 132 Appendix A. STQ- l o o o o o o o o o o o o o B. NATIONAL TENNIS RATING PROGRAM (NTRP) . . . C. COMPETITIVE PSYCHOLOGICAL EFFECTIVENESS SCALE (PIIASE I ) o o o o o o o o o o o D. RATER'S CODING PACKET (PHASE II) . . . . E . STQ-z . . . . . . O . . . . C C F. COMPETITIVE PSYCHOLOGICAL EFFECTIVENESS SCALE: PLAYER'S FORM FOR PHASES II AND III . . . G. COMPETITIVE PSYCHOLOGICAL EFFECTIVENESS SCALE: COACH'S FORM FOR PHASES II AND III . . . . H. DISCRIMINANT FUNCTION ANALYSIS ON 118 STQ-Z STATEMENTS o o o o o o o o o o o o I o STQ-3 o o o o o o o o o o o o o J. LIST OF APPENDICES DISCRIMINANT FUNCTION ANALYSIS ON 50 STQ-3 STATEMENTS . . . . . . . . . . . Page 136 140 143 145 165 172 174 178 181 187 CHAPTER I INTRODUCTION In recent years there has been a growing trend in psychology, including the psychology of sport, to focus on the role that cogni— tions play in mediating behavior change (Bandura, 1978; Borkovec, 1976; Ellis, 1980; Girodo, 1977; Kendall & Hollon, 1981; Mahoney, 1977; Meichenbaum, 1979; Silva, 1980; Suinn, 1980). The term "cogni- tion" is a seemingly simple, but deceptively complex concept, depicting a wide array of thought processes. According to Webster (1975), the global definition of the term cognition refers to "the act or process of knowing including both awareness and judgement" (p. 217). Thus, cognition is a construct which lends itself to a multitude of com- ponents open to investigation. An increasing number of investigators have turned to the field of competitive athletics as a fertile area for research examining the extent of cognitive influences on behavioral outcomes (Bennet & Stothart, 1980; Cratty & Davis, 1984; Feltz, 1984; Kirschenbaum, 1983; Kirschenbaum & Bale, 1984; Long, 1980; Mahoney, 1977, 1984; Martens, 1980; Meyers & Schleser, 1980; Meyers, Schleser, Cooke, & Cuvillier, 1979; Rushall, 1984; Silva, 1980, 1983; Silva & Hardy, 1984; Suinn, 1980; Weinberg, Smith, & Jackson, 1983; Zaichkowsky, 1984). As Mahoney (1979) pointed out: . . . competitive athletics offer many parallels to the demands and stresses of everyday life. As such, they may represent an invaluable microcosm within which we may test and refine our understanding of human adaption. The athlete must learn to cope with a wide range of stressors--performance standards, the experience of failure, aging, and so on (p. 441). The study of the role of cognitions on athletic performance is also important in its own right, since many athletes suffer from debilitating levels of stress in the competitive setting (Smith, 1980). A growing number of cognitiye-behavioral intervention programs have been proposed in an effort to help athletes cope with the stressful situations (Bennet & Stothart, 1980; DeWitt, 1980; Kirschenbaum 8 Bale, 1980, 1984; Lane, 1980; Long, 1980; Martens, 1980; Meyers & Schleser, 1980; Nideffer, 1976; Rotella, Gansneder, Ojala, 8 Billing, 1980; Silva, 1980, 1983; Smith, 1980; Spinelli G Barrios, 1980; Suinn, 1980; Wenz & Strong, 1980). Most of the intervention strategies have been aimed at any one, or combination of the following components: (a) behavior, (b) cognition, (c) environment, and/or (d) physiological responses (Smith, 1980). A number of investigators, however, believe that the cognitive processes are the key component of such strategies (Bandura, 1978; Ellis, 1979; Girodo, 1977; Kendall & Hollon, 1981; Long, 1980; Mahoney, 1977; Silva, 1980; Smith, 1980). Bandura (1978) asserted that most external forces impinging on one's behavior are mediated by cognitive processes. In other words, Cognitive factors partly determine which external events will be observed, how they will be perceived, whether they have any lasting effects, what valence and efficacy they have, and how the information they convey will be organized for future use (Bandura, 1978, p. 345). One component within the cognitive construct, which has recently been receiving more and more attention from psychologists in the clinical, as well as the sport setting, is the phenomenon of self-talk. Self-talk (also referred to as self-statements, inner speech, covert statements, internal dialogue, self-referent speech, automatic thoughts, and internal sentences (Kendall & Hollon, 1981)), has fre- quently been analyzed to determine to what extent it is associated with behavioral outcomes. The study of self-talk and its influence on behavior is by no means a new topic of interest. Buddhists and Hindus, for example, have incorporated the use of internal monologues in an attempt to improve the quality of life, for thousands of years. Over 100 years ago, an Englishman, Francis Galton, examined his own inner speech to see if he could become aware of, record, and analyze thoughts that barely crossed his conscious threshold. In the early 19003, Emile Coué and other psychiatrists associated with the Nancy School of Psychiatry, encouraged their patients to repeat general positive statements to themselves, such as "day by day, in every way. I'm getting better and better." In the 1930's a German psychiatrist, Johannes Schultz, developed autogenic training in which individuals were instructed to repeat phrases to themselves, such as "my left arm is heavy" to help attain deep relaxation. Albert Ellis proposed Rational-Emotive Therapy in the mid-1950's, based on the theory that illogical beliefs were the fundamental cause of maladaptive behavior. During the same time period, Norman Vincent Peale emphasized the "power of positive thinking," and Luria, a Russian psychologist sug- gested that a child's intentional motor behavior was controlled by overt, and eventually by covert, self-verbalizations. These are but a few examples indicative of the interest in the phenomenon of self- talk and its possible link to behavior regulation. Although individuals may not be acutely aware of their moment- to-moment self-talk, researchers have supported the contention that the incidence of automatic self-talk appears to be a common everyday occurrence (Kendall, 1981). This premise can also be supported by casual anecdotal observations. It is not uncommon, for example, to observe peOple talking out loud to themselves in private and/or public places. These individuals often seem to be so deep in thought that they appear oblivious to most external distractions. Individuals also regularly engage in covert self-talk as they, for example, decide what they are going to wear for the day; what they are going to eat (or should not eat); how they are going to schedule their day, etc. Evidence of athletes engaging in overt and/or covert self—talk is just as prevalent. In the tennis setting, for instance, John McEnroe and Martina Navratilova are classic examples of athletes who engage in overt self-talk while competing. A variety of sports related investigations have documented that athletes do talk to themselves, while competing (Gould, Weiss, 8 Weinberg, 1981; Klinger, Barta, 8 Glas, 1981; Mahoney 8 Avener, 1977; Silva, 1980; Van Noord, 1980)- Even though the phenomenon self-talk has been sporadically addressed for centuries and although it may occur on a daily basis (Kendall 8 Hollon, 1981), extensive empirical research on the possible connection between an individual's inner speech and subsequent behavior has begun to mushroom only within the past decade. Much of the research completed in the clinical and sport setting has focused on the altera- tion of subjects' self-statements to mediate a behavior change (Beck, 1979; Bistline, Jaremko 8 Sobleman, 1980; Ellis, 1979; Girodo, 1977; Gravel, Lemieux, 8 Ladouceur, 1980; Hamberger 8 Lohr, 1980; Long, 1980; Meichenbaum, 1979; Meyers 8 Schleser, 1980; Novaco, 1976; Silva, 1980; Suinn, 1980). The significant results of many of the cognitive-behavioral studies have been interpreted as evidence that self-statement modifi— cations were responsible for treatment success. These may be prema- ture assumptions, however, since the majority of investigators did not incorporate valid and reliable self-statement assessment measures to document the asserted changes. It seems inappropriate, therefore, in the absence of such documentation, to proclaim that self-statement modification was solely responsible for treatment success (Class, Merluzzi, Biever, 8 Larsen, 1982). Hence, it has yet to be determined to what extent self-statement modification influences behavioral out- comes. Just as behavioral assessments were needed for the growth and validation of behavioral interventions, self-statement assessments are needed for the development and expansion of cognitive interventions (Glass et al., 1982). Investigators have begun to develop in-depth self-statement assessment instruments only within the past few years. To date, these assessment instruments have focused on self—talk related to depres- sion (Hollon 8 Kendall, 1980), social anxiety (Glass et al., 1982), athletic performance (Klinger, Barta, 8 Glas, 1981), surgical coping responses (Kendall, Williams, Pechacek, Graham, Shisslak, 8 Herzoff, 1979), assertiveness (Schwartz 8 Gottman, 1976), test anxiety (Hollands— worth, Glazeski, Kirkland, Jones, 8 Van Norman, 1979), and acute anxiety (Hurlburt 8 Sipprelle, 1978). The Klinger et a1. study is the only published study to date which incorporated a detailed assess- ment on athletes' competition-related self-talk. Thus, the recent proliferation of cognitive intervention programs designed to improve athletic performance (Gravel et al., 1980; Long, 1980; Martens, 1980; Meyers 8 Schleser, 1980; Silva, 1980; Smith, 1980; Suinn, 1980) have been develOped in the absence of sound empirical evidence. Therefore, there is an urgent need to develop valid and reliable assessment tools to: (1) determine the content of an athlete's game-related self-talk; (2) develop theories regarding the relationship between self—talk and subsequent performance; and (3) establish instruments which will provide researchers with information about the quantity and quality of athletes' cognitive changes. This investigation, then, focused on the development and validation of an assessment tool to examine the competition-related automatic self—talk of a select group of tennis players. Tennis players were solicited to participate in this investiga- tion for a number of reasons. First, tennis is generally thought of as a sport in which psychological factors play an important role (Gallwey, 1974; Gologor, 1979; Kauss, 1980; Spinelli 8 Barrios, 1980). Second, since tennis is an individual sport, the relationship between competition-related self-talk and performance outcomes should be minimally confounded. Third, tennis is a popular lifetime sport, played by both males and females, of all ages and levels of skill. Thus, the implications of this study and follow-up studies may be applicable to a large number of individuals. Statement of the Problem The purpose of this investigation was to develop a self-talk questionnaire (STQé3) which could validly and reliably assess the type and frequency of tennis players' competition-related self-talk as it related to competitive psychological effectiveness. The resultant STQ-3 was then used to test the following hypothesis: Hypothesis: Two criterion groups of psychologically effective and ineffective tennis players will significantly differ in their competition-related self-talk responses, as measured by the STQ-3. Significance of the Study The results from a wide array of both clinical and sports- related investigations, seem to suggest that self-talk and behavior, and hence an athlete's performance, are interrelated. Even though a number of researchers have already employed intervention programs which have contained a self-talk modification component to enhance player performance (Gravel et al., 1980; Long, 1980; Meyers 8 Schlesser, 1980; Silva, 1980; Smith, 1980; Suinn, 1980), the extent to which self-talk is related to performance outcomes has yet to be determined. Before additional intervention programs are implemented to examine the relationship between self-talk and performance, it is necessary to determine what constitutes an athlete's competition related self—talk. That is, basic principles regarding the nature of self-talk are needed to guide subsequent empirical investigations. In order to assess an athlete's game-related self-talk, researchers must rely on valid and reliable self-reports. Although self-reports may not be all inclusive and may be affected by environmental factors (Genest 8 Turk, 1981), they, nevertheless, are themselves behaviors which can serve as data to confirm or disconfirm research hypotheses (Kendall 8 Hollon, 1981). Hence, it is imperative that empirically sound self-statement assessment tools he developed to give researchers reliable and valid data with which to build their theories. The major purpose of this investigation was to develop an explora— tory self-talk assessment tool, which could validly and reliably assess the competition—related self—talk of tennis players as it pertained to competitive psychological effectiveness. Additionally, it was hoped that the STQ-3 could assist in the following assessment functions as outlined by Nideffer (1976). 1. An aid to assist in the identification of possible self-talk problem areas 2. An aid to increase the coach's understanding of player behavior 3. An aid to improve communication channels between both the coach and the athlete In summary, this investigation was intended to serve as the first step in a series of exploratory studies to systematically examine the self-talk content emitted by tennis players. It was hoped that this investigation would stimulate future research to empirically explore the interrelationships between an athlete's competition-related self-talk, and corresponding performance patterns. Assumptions Underlying the Study Basic to this investigation were the following assumptions: 1. Tennis players engaged in self-talk while competing. 2. Tennis players, when given specific match situations to focus on, could recall their game-related self-statements from a tennis match played within one week prior to recall (Kendall 8 Hollon, 1981). 3. Tennis players could distinguish, by means of 5-point scale, the relative frequency with which they emitted a particular self-statement. 4. The situations which were proposed to elicit self- statements in Phase I were common situations experienced by the majority of tennis players. 5. The pool of self-statements collected for Phase I was representative of the types of statements tennis players said to themselves while competing. 10 6. Tennis players with a 3.0 NTRP rating or higher, selected from Florida, Michigan, and Texas, comprised a relatively homogeneous sample. 7. Coaches were able to accurately assess and subsequently rate each of their players' overall competitive psycho- logical effectiveness, based on a 5-point scale. Scope of the Study This investigation was limited by the following boundaries: 1. The self-talk analyses and conclusions were limited to the final STQ-3 items. 2. The STQ analyses were limited to the responses from 302 participants from Texas, Michigan, and Florida. 3. The phenomenon of Competitive Psychological Effective- ness was bounded by the properties of a 5-point scale. 4. The accuracy of the self—talk item responses were limited to the preciseness of subject self-talk recall and unbiased recordings. 5. The accuracy of the player skill ratings was dependent upon each coach's interpretation of the NTRP scale. 6. Data for the development of the STQ were collected during 1981—82. Definition of Terms The following terms were defined according to their use in this investigation. 11 Competitive Psychological Effectiveness (CPE): A rating, based on a 5-point scale, which indicated the degree to which an athlete was steady, consistent, and reliable under the pressure of a competi- tive event . Egagh: The individual who was in charge of organizing and con- ducting practices, ranking his/her players based on skill and match performance, and accompanying his/her team when they competed against players from the Opposing teams. National Tennis Rating Program (NTRP): The NTRP was adopted in 1979 by the National Tennis Association, the United States Tennis Association, and the United States Professional Tennis Association, as a unified method to classify tennis players of all skill levels across the country. The NTRP consists of a 13-point scale which describes the skill levels of a rank beginner, at one end of the scale, to a world class player, at the other end. Coaches and teachers who are members of any one oftfimepreceeding tennis associa- tions automatically receive a packet of material which gives explicit instructions on how to rate their players according to the NTRP. Additionally, all coaches and teachers who participated in this investigation were given a copy of the NTRP scale along with the STQ packet. Psychologically Effective Player: A tennis player who received a rating of 4 or 5 from his/her coach/teacher on the Competitive Psy- chological Effectiveness Scale. This type of player played to 12 the top of her/his porential during a match anywhere from 60 to 100% of the time, and was often considered to be a steady and consistent player. The psychologically effective player often won when competing against an opponent with a comparable skill level. Psychologically Ineffective Player: A tennis player who received a rating of 1 or 2 from her/his coach/teacher on the Competitive Psy- chological Effectiveness Scale. This type of player played to the top of her/his potential during a match anywhere from 0 to 20% of the time, and was often considered to be an unsteady and inconsistent player. The psychologically ineffective player often lost when com- peting against an opponent with a comparable skill level. Self-Statements or Self-Talk (competition-related): Words or phrases players said to themselves, either overtly or covertly, while competing. This included all self-instructions; task relevant or irrelevant statements; affective statements such as disgust, pleasure, frustration, anger; observations; self-criticisms; self-reinforcement, etc. Self—Talk Questionnaire (STQrB): The assessment instrument which was developed during this investigation to examine the self-statements emitted by tennis players. Tennis Competition: A tennis match which had been assigned by a coach, teacher, or tournament draw. Tennis Class/Team: A high school, college, or club-sponsored tennis organization which was led and organized by a tennis teacher/ coach, whose job was to lead practices, aid members in skill develop- ment, and arrange competition. CHAPTER II REVIEW OF RELATED LITERATURE AND RESEARCH Influence of Self—Talk on Behavior The renewed vigor with which contemporary research on self-talk has been conducted has been promoted in part (1) by research focus- ing on children with learning deficiencies which showed that their behavior could be altered by the regulation of their overt self- verbalizations (Kendall, 1977; Meichenbaum, 1979; Meichenbaum 8 Goodman, 1971; Robin, Armel, 8 O'Leary, 1975); and (2) by cognitive- behavioral theories proposed by cognitive therapists such as Albert Ellis (1979) and Aaron Beck (1979). Contemporary researchers who have examined the relationship between self-talk alteration and subsequent behavior change seem to have produced a pool of divergent results. There are a number of possible explanations which could account for the incompatible find- ings, including differences in: (1) age of subjects, (2) type of experimental settings, (3) type of intervention programs, and (4) dimensions of self-talk. Each of these factors will be discussed in the following sections. 13 14 Age of Subjects 'The contemporary surge of self-talk research was spurred in part by investigations which focused on the regulation of the overt speech of children with learning deficiencies to improve their task perform- ance. However, a number of problems occur when the results of such studies are used to support the global notion that an alteration of an individual's self-talk will lead to a behavior change. First, the cognitive functions of children and adults may differ due to develop- mental processes (Wadsworth, 1971). Therefore, the cognitive- behavioral theories drawn from research on children may not be directly applicable to adults. Second, various researchers have used the results of studies which incorporated an examination and modifica- tion of childrenis 21335 self—talk to build research hypotheses designed to examine and modify adults' covert self-talk. Although there may be many similarities between the task-related overt self- talk of children and the covert self-talk of adults, it is just as likely that there may also be a number of significant differences. It may be an erroneous assumption, therefore, to conclude that precisely the same principles apply to both overt and covert self-statements (Kendall 8 Hollon, 1981). Third, a great deal of the cognitive- behavioral research completed on children has focused on improving the test-taking skills of children who displayed impulsive behaviors. Frequently, these children were given paper and pencil tests and were taught to verbalize a variety of problem-solving self-statements to improve their performance as they took a test. Much of the 15 cognitive-behavioral research completed on adults, on the other hand, has focused on methods of coping with a variety of social stressors, such as anxiety (Bistline, Jaremko, 8 Sobleman, 1980; Bruch, 1978; Cooley 8 Spiegler, 1980), phobias (Meichenbaum, 1979), fear (Girodo 8 Roehl, 1978; Norton, Maclean, 8 Wachna, 1978), and anger (Hamberger 8 Lohr, 1980; Novoco, 1976). It is possible, therefore, that the intervention programs devised to enhance the test-taking skills of impulsive children may not be totally applicable to those interven- tion programs designed to improve the stress-related coping responses of adults. Hence, extreme care must be taken not to blindly generalize the results of research to individuals or situations which were not the original focus of the investigation. Type of Experimental Setting Since laboratory and field settings are inherently different, it is possible that differential findings regarding the influence of cognitions on behavior may have been due, in part, to the effects of a variety of confounding variables indigenous to the specific experi- mental setting. For example, the effects of social facilitation may have confounded the findings of some of the laboratory-based experi- ments. That is, A person's motivation to comply with the demands of the experiment may make him [sicl feel more obligated to go along with the use of self-instruction throughout his [sic] perform- ance in the experiment. If these motivational factors interact with attention diversion components of cognitive behavior, we can wonder to what extent we might expect these techniques to be successful outside the laboratory where people may not be so compliant (Girodo, 1977, p. 238). 16 Differences in research findings among laboratory and field investigations may also be partially due to variations of external and internal validity. That is, laboratory investigations generally have high internal validity and low external validity, whereas field experiments often have low internal and high external validity (Martens, 1975). Thus, caution should be used when comparing results and making inferences from research completed in the laboratory setting to the field setting, or vice versa. Type of Intervention Program The use of eclectic self-talk intervention programs to modify behavior may also have contributed to the conflicting results reported in the cognitive behavioral literature. The major problem with such intervention programs is the difficulty in determining the efficacy of each of the component parts. Meichenbaum (1979), one of the leading theorists in the contem- porary cognitive-behavior modification movement, developed an eclectic Stress Inoculation Training (SIT) program which included components which focused on the assessment and modification of maladaptive self- statements. His program incorporated educational rehearsal and application training phases, each consisting of a variety of inter— vention techniques. The educational training phase included informing clients that their anxiety reaction consisted of increased arousal coupled with inappropriate self-statements which promoted the mala- daptive behavior. The rehearsal phase consisted of teaching the client a variety of behavioral coping responses, as well as cognitive 17 coping responses which included monitoring and modifying negative and anxiety-provoking self-statements. During the application phase the clients learned to incorporate their caping skills when faced with a variety of stressors. The entire treatment package was successful in aiding individuals to cope with interpersonal anxiety, enhance anger control, increase pain tolerance, and control phobic reactions. However, since the components of the SIT program were not separated and systematically analyzed, the precise relation- ship between self—talk modification and subsequent behavior change was not determined. Girodo and Roehl (1978) designed a study to test the efficacy of the SIT program. The investigators focused on the coping responses of female subjects who had expressed a fear of flying. The women were divided into one of the following four groups: self-statement train- ing (SIT Procedure), preparatory information training, a combination of self-statement and preparatory information training and control. Girodo and Roehl found that subjects in the self-statement training group who engaged in positive self-talk during the stress of an "unexpected" missed landing, did not significantly reduce their anxiety. The investigators hypothesized that the preoccupation of the subjects with their memorized list of statements may have prevented them from utilizing effective coping strategies during the missed landing. These results re—emphasize the point that eclectic intervention pro- grams such as the SIT, which work in one situation, may not neces- sarily be effective in another. 18 In a sports-related investigation, Meyers and Schleser (1980) also incorporated a self-statement modification component as a part of an eclectic intervention program to improve the performance of an athlete in the competitive setting. In an uncontrolled case study, the investigators assessed and then treated a 22-year-old male inter- collegiate basketball player who was exhibiting concentration problems in the game situation. The intervention program included such com- ponents as: identification of attentional problems, progressive relaxation, cognitive intervention rationale, coping orientation including thought-stopping and self-instructions imagery, and in vivo practice. A post-intervention analysis of his game performance indi- cated that the athlete's percentage of successful field goals signifi- cantly increased. Even though the results of this investigation are encouraging, it should be emphasized that this was an uncontrolled case study, and that no attempt was made to examine the relative effectiveness of each of the component parts. Therefore, judgment on the efficacy of the overall treatment program, as well as on each of the intervention procedures, should be reserved until further research is completed to clear up the present uncertainties. An equally successful, but empirically weak, eclectic interven- tion procedure designed to enhance an athlete's performance, was proposed by Silva (1980). His intervention program incorporated three major phases: an identification phase, during which the athlete's problem was defined in detail; a cognitive restructuring phase, during which the athlete was persuaded to replace a particular pattern of unproductive self-statements with a more appropriate coping 19 chain of self-statements; and a pairing phase during which the athlete was taught to pair a key word or phrase with a specific image of the problematic competitive situation. Silva utilized his intervention program in a series of three single case study investigations. Two male intercollegiate basketball players and one male intercollegiate hockey player volunteered as subjects. Each subject was experiencing performance difficulties in his respective sport prior to participa- tion in the intervention program. Results indicated marked improve- ment for each athlete according to pretest to posttest game sta- tistics. Again, while these results are encouraging, only tentative support should be given to Silva's cognitive intervention program, due to the multiphasic case study design and lack of control. The majority of researchers who have prOposed eclectic cognitive- behavioral intervention programs did not initiate a subsequent com- ponent analysis to determine the relative importance of each of the contributing components. Furthermore, the results from the use of multiphasic programs are often difficult to compare, since the com- ponents of each program vary depending on a number of factors includ— ing the type of behavior to be altered and the research interests of the investigator. At this point in time, the precise operational mechanisms of the various self-talk components incorporated in these programs have not yet been determined. Therefore, much research is still needed to determine why a self-talk component may be effective under certain conditions (Meichenbaum, 1979), and not others (Girodo 8 Roehl, 1978). Once the underlying mechanisms of each contributing 20 component are understood, it will be possible to streamline the existing multiphasic cognitive—behavioral programs into efficient and effective systems. Dimensions of Self-Talk Self-talk is a complex phenomenon. It seems that it varies at the very least, along the following two dimensions: overt to covert, and.induced to automatic. Figure 1 depicts a schematic representa- tion of these two basic dimensions. Overt Induced Automatic (naturally occurring) Covert Figure 1. Schematic Representation of Two Basic Self-talk Dimensions. An individual's inner speech can also vary along a subconscious to conscious continuum. Disagreement and confusion about the effects of self-talk may have been promoted in part, therefore, by researchers 21 who did not ferret out the different types of self—talk and have tried to generalize or equate research findings which focused on one type of self-talk to another. As indicated previously, such generaliza- tions have been made in the attempt to equate children's overt self— talk with the covert self-talk of adults. Similarly, it is equally confusing if investigators fail to differentiate between the effects of induced and automatic self-talk. For example, subjects are often requested to repeat a prearranged, memorized list of self-statements to themselves to replace their maladaptive self-talk in a given situation (Girodo, 1977; Girodo 8 Roehl, 1978; Girodo 8 Wood, 1979; Meichenbaum, 1979; Meyers, Schleser, Cooke, 8 Cuvillier, 1979). This induced self-talk, however, may be much different from the automatic inner speech used by an individual when confronted with a similar situation, outside of the experimental setting. Girodo (1977) has proposed that at least three interrelated cognitive mechanisms may be operating when an individual who is faced with a stressor is induced to emit coping self-statements: self- persuasion, attention diversion, and cognitively cued behavior. He suggested that unless subjects could be persuaded that their particu- lar stressor was actually nonthreatening, the act of emitting self- statements would simply act as an attention-diversion device. The hypothesis which suggests that the act of emitting induced self-statements might direct attention from the task at hand, may have received at least partial support from the investigations by Meyers et a1. (1979). They conducted two studies to examine the effect of induced self-instruction on the development of various gymnastic skills. 22 The investigators found that the act of emitting a set of prearranged self-instructions did not significantly improve the gymnasts' per- formance over that of the other experimental groups. It is possible that the induced self-talk may have diverted the subjects' attention away from the task due to a number of methodological problems. Meyers et a1. chose 40 beginning female gymnasts, ages 7 to 17, to serve as subjects for their first investigation. The gymnasts were pretested on their ability to perform a demonstrated skill and were then stratified into onecflffour self-instructional groups: positive, negative, coping, or neutral self-instruction. Subjects in the posi- tive, negative, and cOping self-instructional groups initially were read the entire set of self-instructions, which was approximately 200 words long. At the end of each sentence or logical break, they were required to repeat the instructions back to the instructor. Next, this entire process was repeated twice with both the subjects and instructor whispering the instructions back and forth, and finally, two more times as the subjects were instructed to think about the instructions. These subjects then had to repeat the set of instructions to themselves as they performed a gymnastic skill 20 times. Subjects in the neutral self-instruction group received task instruction, but not self- instructional techniques. These subjects were also required to per- form the skill 20 times, but were not required to engage in induced self-talk. The entire experimental process was repeated for a second, more difficult gymnastic skill. Although subjects in all four groups showed significant pre- to posttraining improvement, the investigators 23 found no significant differences between the groups on either task. In their second investigation, Meyers et a1. (1979) examined the effect of three experimental conditions: physical practice, cognitive practice, and combined physical and cognitive practice on gymnasts' performance. The subjects were 36 females, ages 7 to 15, enrolled in a beginning gymnastics class. Those in the physical practice group were allowed a total of 20 minutes to practice the specified task. Subjects in the cognitive practice group were read a positive set of self-statements, similar to those used in the pre- vious experiment. The self-statements were read to the subjects once for each of two 10 minute cognitive practice periods, after which the subjects were asked to rehearse the self-instructions on their own for the remainder of each 10 minute period. Subjects in the combined cognitive and physical practice group spent a total of 10 minutes physically practicing the task and 10 minutes cognitively rehearsing the task. The entire experimental procedure was repeated three times for a low-, moderate-, and high-difficulty task. Meyers et al. found that the physical practice group performed significantly better on the low- and high-—difficulty task than the cognitive practice group. Furthermore, they found that there were no significant differences between the cognitive practice group and the no-practice control group. All groups significantly improved their performance, with the exception of the cognitive practice group on the low-difficulty task and the no practice control group on the high-difficulty task. However, there 24 was a nonsignificant trend which indicated that the combination of cognitive and physical practice actually seemed to reduce the effec- tiveness of the physical practice. There may be several reasons for the apparent lack of cognitive influence on motor skill development and improvement in the Meyers et al. investigations. First, the mean ages of the subjects in the two studies were 10.1 and 11.0 years, respectively. The rather lengthy set of instructions may have been too complex for children. Consequently, it is possible that the gymnasts' performance would not have significantly improved, regardless of the type of self—instructions that had been emitted. The subjects may have experienced infor- mation overload, which, in turn, may have interfered with maximal learning. Second, even if the set of specific self-instructions had been shorter, the study may have produced similar results. That is, the induced and the naturally occurring self-talk may have had dis- similar effects on motor performance. Meyers et al. made no attempt to distinguish between these two types of self-talk. In fact, their investigations were promoted by an exploratory survey study (Mahoney 8 Avener, 1977) in which the results indicated that the naturally occurring self-talk of male Olympic gymnasts, who qualified for the United States team, was more of a coping and task-relevant nature than the nonqualifiers. Hence, Meyers et al. had hypothesized that appropriate induced self-instructions might enhance the development of various gymnastic skills. Consequently, when they had to conclude that the performances of the induced self-talk groups were not 25 significantly superior to the other groups, they questioned the contribution of self-talk to the development and enhancement of motor skills. Until additional research focuses on the similarities and differences between induced and naturally occurring self-talk, care must be taken not to generalize the findings from research on automatic self-talk to situations employing the use of induced self- talk, and vice versa. Conclusions Before accurate comparisons can be made, however, on the simi- larities and differences between induced and automatic self-statements, separate in-depth analyses are needed to assess both types of self— talk. Furthermore, it would seem that researchers should focus empirical analyses on the make-up of an individual's naturally occurring self-talk before attempting to implement induced self-talk procedures. Theorists have already made an array of assumptions regarding what constitutes automatic self—talk, and its role in govern- ing behavior. These self-talk theories, however, have often been based on weak empirical studies, or at times, even based on past clinical experience. For example, Ellis (1979), the founder of rational-emotive therapy was convinced that behavior and emotions were governed by what individuals said to themselves. Spinelli 8 Barrios (1980) have contended that individuals who emit negative or distracting self-statements, divert their attention away from the task at hand. Lanyon 8 Lanyon (1978) suggested that if individuals were encouraged to focus on their self-talk, they would be able to recognize 26 the cognitions that act as cues for problem behaviors. Smith (1980) who deveIOped a multifaceted Stress Management Training program for athletes, stated: ". . . athletes can be trained to discover, chal- lenge, and change the internal self-statements that are, in actuality, generating their stress responses" (p. 58). Thus, even with no solid proof of the extent of the correlation between self-talk and behavior, some investigators have directly stated or inferred that there is a causal relationship. It seems that many researchers have modified and/or have promoted the use of induced self-talk in their empirical investigations, before establishing basic principles on the nature of automatic self-talk. As has been shown in the foregoing discussion, the intent of much of the research in this area has been to demonstrate a causal link between self-talk and subsequent behavior. Thus, self-talk has often been analyzed as an independent variable. However, in order to build and test theories on its effects, it is also necessary to study the phenomenon of self-talk as a dependent variable. Its content, frequency, and occurrence, therefore, need to be investigated in detail. Hence, this investigation attempted to assess the naturally occurring self-talk of subjects, without making any cognitive and/or behavioral manipulations. This investigation focused entirely on the development of an instrument to measure self-talk. Assessment of Self-Talk A few investigations have been designed to include an analysis of the occurrence of athletes' competition-related self-talk. In one 27 of the first studies to include an examination of an athlete's self- talk. Mahoney and Avener (1977) administered a multi-purpose question- naire to 12 elite male gymnasts 48 hours before they participated in the 1976 American Olympic qualification trials. Only a small portion of the questionnaire dealt with self-talk. These few questions focused on self-statement occurrence and frequency, rather than specifically on what the athletes said to themselves. Results indicated that the gymnasts who qualified for the 1976 Olympic team engaged in more self-talk during training and competition than did the nonqualifiers. Based on the results of the post-event interviews with the athletes, the investigators surmised that the gymnasts who made the team engaged in appropriate coping self-talk more often than did the nonqualifiers. A follow-up investigation was conducted by Meyers, Cooke, Cullen, and Liles (1979). Nine highly skilled members of the Memphis State University men's racquetball team served as subjects. These athletes were administered a modified version of the Mahoney and Avener (1977) questionnaire at two different points during their competitive season. Prior to both administrations of the questionnaire the coach ranked each of the players according to skill. Again, the few questions which focused on self-talk were designed to assess the occurrence and fre- quency of the athletes' self-statements. The results were similar to those found by Mahoney and Avener (1977) in that the higher a player's skill ranking, the more he tended to engage in self-talk in both train- ing and competitive situations. 28 Highlen and Bennett (1980) incorporated an adapted version of the Mahoney and Avener (1977) questionnaire to examine the behavioral and cognitive strategies used by elite wrestlers. Forty outstanding Canadian male wrestlers, who were competing for positions on national teams, were selected as subjects. The wrestlers ranged in age from 18 to 32. A total of 16 out of the 40 wrestlers did not qualify for a national team. In contrast to the findings by Mahoney and Avener, and by Meyers et al. results in this study indicated that there were no significant differences in self-talk frequencies between the quali- fiers and nonqualifiers for the Canadian national teams. Furthermore, Highlen and Bennett found that on the average, wrestlers engaged in only a moderate degree of self-talk. Gould, Weiss, and Weinberg (1981) incorporated a similar multi- faceted questionnaire to assess the cognitive and behavioral strategies used by collegiate wrestlers. Forty-nine out of 99 male intercollegiate wrestlers participating in the 1980 Big Ten Championships volunteered to answer the questionnaire within 24 hours of the tournament. Since an in-depth self-talk analysis per se, was once again not the intent of the questionnaire, only a few questions focused on the occurrence and frequency of the athletes' self-talk. Analysis of the responses from the entire sample of subjects indicated that most wrestlers fre- quently engaged in self-talk during training and competition. In fact, the mean frequency was 7.75 based on an 11-point Likert scale. How- ever, similar to the findings by Highlen and Bennett, there was no significant difference between the frequency of self-talk by place winners and those who did not place. 29 Based on the findings of these studies, it appears as though many athletes may talk to themselves in competitive situations. Assum- ing that this is true, the next step is to analyze the content of the self-statements. Only one sports-related study to date was specifi- cally designed to assess the automatic self-talk of athletes during competition (Klinger, Barta, 8 Glas, 1981). Outside of the athletic setting, only a handful of clinical self-talk assessment investiga- tions have been completed.' Yet, investigators have already incor- porated a variety of assessment methods. Based on a review of available literature, Kendall and Hollon (1981) divided the methods of assessment which have been employed into the following four categories: (1) record- ing, (2) sampling, (3) production, and (4) endorsement. These four methods are, however, neither all inclusive or exclusive. Thus, there is a certain amount of procedural overlap among them. Nevertheless, they provide a practical way to examine the majority of the existing self-statement assessment methods. These four categories will each be briefly discussed. Recording Investigators who have employed recording methods have incor— porated the use of audio and/or audio-visual equipment to aid in self-talk assessment. Audio equipment has been frequently used to record and subsequently assess the ongoing overt self-talk of indi- viduals while they performed a specified task. The use of audio- video equipment has assisted subjects in the recall of their task- related self-talk by enabling them to view a replay of their performance. 30 Kendall and Finch (1979) incorporated the use of a tape recorder to analyze the verbal behavior of impulsive (treated and untreated) and reflective children, while they worked on the Matching Familiar Figures test. The investigators coded childrens' verbal reSponses into the following six categories: task-related questions, off—task verbalizations, statements of task difficulty, thinking outloud, verbalizing the answer, and total on-task verbal behavior. Results indicated that the impulsive children who were treated, engaged in significantly more total on-task verbalizations than the untreated impulsive, and,than the reflective children. Furthermore, the treated impulsive children engaged in less off-task verbalizations as compared to the reflective children. Van Noord (1980) utilized the audio-video playback technique to examine the self-talk of volleyball players. The exploratory and uncontrolled investigation was designed to see if volleyball players engaged in self-talk while competing, and if so, if they could retro- spectively recall and record their game-related self-statements. Four out of ten varsity players from the 1980 Michigan State University women's volleyball team volunteered to participate in the study. Three different game conditions were audio-video taped: practice scrimmage, dual meet, and state tournament semi-final match. Imme- diately following each of the game conditions, subjects were asked to watch and listen to the audio-video replay of their preformance. The tape was randomly stopped on the average of 11 times per game condition, 31 at which time the players were asked to recall and record what they said to themselves (if anything) before, during, and after that par- ticular game point. Results indicated that the athletes experienced little or no recall problems. Furthermore, the four athletes reported self-talk during all three time periods (before, during, and after a point) for each of the three game conditions. Most of their self- talk responses prior to and during the skill consisted of self- directions and/or a description of the game situation. Self-talk after the skill was too diverse to discern any one prevalent type of self-statement. Both audio and audio-video equipment were also utilized in an investigation which examined the relationship between subjects' self- statements and test anxiety (Hollandsworth, Glazeski, Kirkland, Jones 8 Van Norman, 1979). After taking a written test, subjects reviewed a videotape of themselves taking the test and were asked to say whatever they were feeling and/or thinking at 13 preselected points. Subjects could also stop the tape at will if they wanted to make a comment. All statements were audio-taped and categorized as on-task, off-task, positive, negative, or miscellaneous. Results indicated that test- anxious subjects tended to engage in more negative and off-task thoughts than the low text-anxious subjects. Even though use of the recording methods have not been empiri- cally validated, the procedures appear to have certain advantages. The recording process could possibly aid subjects in their self- statement recall and/or enable investigators to audio-tape and subse- quently assess subjects spontaneous overt self-talk. The recording 32 procedures, however, may also have a few inherent weaknesses. First, investigators such as Kendall and Finch (1979) who recorded subjects' overt self-talk, have often neglected to do parallel analyses of the subjects' corresponding covert self-talk. This omission would make it difficult to determine whether a subject's silence indicated that she/ he was actually silent or engaging in covert self-talk. Furthermore, this omission would also prevent the researcher from determining if the individual's overt self-talk was indicative of his/her covert self- talk. It is conceivable, for example, that an individual could emit on-task overt self-statements and off-task covert self-statements. The second weakness of the recording method is that subjects, especially adults, may censor their overt self-statements as a reac- tive effect. The extent to which this has occurred in studies employ- ing this technique has not yet been determined. Furthermore, when subjects were instructed to "think out loud" while performing a task, their overt verbalizations may have not accurately depicted their natural, automatic self-talk. For example, Pope (as cited in Kendall 8 Hollon, 1981, p. 94) found that individuals who were asked to "think out loud" took longer to emit certain thoughts than subjects who engaged in covert self-talk. The subjects who were instructed to "think out loud" may have felt obligated to complete their sentences and/or to censor their thoughts. It is possible than that their per— formance on a task while "thinking out loud" was not an accurate reflection of their "natural" task performance. Thus, the use of some of the recording methods reviewed may have been based on the assump- tion that overt and covert self-talk were synonymous. 33 Sampling A second technique which has been used to assess self-talk is the sampling method. The sampling method refers to a technique in which subjects are cued at random time intervals, and then record self-statements emitted at that specific moment. The purpose of this procedure is to obtain an unbiased estimate of the frequency with which various self-statements are emitted. For example, Hurlburt and Sipprelli (1978) conducted an experi- ment on a male subject who was suffering from severe anxiety attacks. The investigators instructed the man to wear a small battery-operated signal generator, which emitted tones at random time intervals. As soon as the subject heard the tone, he was instructed to tune into his self—talkanzthat moment, and to record his self—statements. After the investigators tabulated the subject's self¥statements, they dis- covered that the highest frequency of statements were related to a family problem. Once the source of the subject's anxiety attacks was pinpointed, subsequent therapy sessions were successful in mini- mizing his anxiety attacks. Only one study (Klinger et al., 1981) has employed the sampling method in the athletic setting. This study will be discussed in detail, since it is the only published investigation to date which was specifically designed to assess the automatic self-talk of athletes in a competitive setting. The investigators utilized 14 members of a small college basketball team (gender not specified) to examine the relationship between the self-talk of athletes and concurrent team 34 momentum. Players narrated their thoughts into a tape recorder at various points during the course of four intercollegiate games, two which the players won and two in which they lost. Players were told to recall their most recent self-statements as soon as the coach sig- naled for their substitution, and to narrate their thoughts and mood into a tape recorder immediately after walking off the court. Players sitting on the bench were also randomly asked to narrate their current self-statements and mood into the tape recorder. Players were asked to recall their self-talk on the average of one to two times per game. This yielded a total of 90 analyzable self-statements. After examining the self-statements, the investigators decided to code the statements innoone of the five following categories: (1) com- pletely unrelated, (2) irrelevant to good play, (3) difficulty/ ability, (4) self-exhortation[evaluation, and (5) problem-solving/ perceptual. Approximately one-fourth of the statements were coded as "problem-solving[perceptual," 2% were categorized as "completely unrelated," while the remaining three categories each contained approxi- mately one-fifth of the statements. Of the statements, 14% were designated as '"unratable. " The frequency of statements in each of these categories was then compared to the team's corresponding game momentum. Investigators defined "hot times" as periods when the team scored two or more consecutive field goals against the opponents. "Gap times" were defined as periods when the opponents scored two or more consecutive field goals. The remaining game time was defined as "other." The 35 largest percentage (53%) of statements collected during "hot" times were categorized as "problem-solving/perceptual," whereas no such statements occurred during "gap" times. Conversely, half of the statements collected during "gap" times were classified as "self- exhortation/evaluation" while no such statements occurred during "hot" times. The investigators also found that the relationship between the type of self-talk emitted and total number of points scored by an athlete was inconsistent. Furthermore, there appeared to be no cate- gorical differences between the self-statements recorded during halves/games won or lost. On the basis of their results, Klinger et al. (1981) suggested that perhaps a concentration shift occurred during "gap" time, when players focused on themselves, rather than the task at hand. Perhaps this apparent concentration shift was related to a similar phenomena reported by Nideffer (1976) who found that anxious athletes tended to direct their attention inward and became overloaded by self-statements focused on themselves. Although the data of the Klinger et al. (1981) study indicated a possible relationship between type of player self-talk and corre- sponding team performance, the results of this investigation should be viewed with caution for a number of reasons. First, the sample size was small and only one team was observed. Second, the vast majority (86%) of the recorded self-statements came from the players sitting on the bench. That is, only 13 of all self-statements recorded came from players coming off the court after a substitution. Further- more, since each athlete was thought sampled only one to two times per 36 game, a total of approximately three self-statements per game were gathered from players coming off the court. This may partially explain the inconsistent relationship between a player's self-talk and points scored. Third, the analysis did not focus on individual players, but rather treated the entire team as a single unit. The success of the sampling method in assessing automatic self- talk may depend, in part, on the setting in which the assessment and subsequent recording takes place. That is, it may be impractical in many types of natural settings to randomly stop all activity to recall and record self-talk at that moment. In most sport settings, for example, it may be very difficult and/or impractical to obtain on- the-spot reports of an athlete's self-talk during a competitive event. Nevertheless, the sampling method does have the advantage of limiting the occurrence of selective reporting. Generally speaking, however, this method of assessment may be more suitable for certain clinical populations than for athletes in the competitive setting. Production A third mode of self-statement assessment incorporated by researchers has been the production method. This procedure requires subjects, after being exposed to a particular situation to recall and record their self-statements, immediately after being exposed to a particular situation. One of the most common techniques incorporated in this method is thought listing. For example, Cacioppo, Glass, and Merluzzi (1979) utilized this technique in their examination of hetero- social anxiety. High social anxious and low social anxious male 37 undergraduates were led to believe that they would be engaging in a discussion with a female undergraduate, either via an intercom system (high anonymity condition) or face-to-face (low anonymity condition). After the subject was told what to expect, the experimenter left the room for three minutes to give the subject time to reflect on the upcoming discussion. When the experimenter returned, the subjects were asked to list all the thoughts that went through their head during the three-minute wait in anticipation of the interaction with the female. In addition, subjects were asked to rate each of their thoughts as positive, neutral, or negative. Results indicated that high socially anxious males engaged in more negative self-talk than the low socially anxious males. A major strength of the production method seems to be that it provides the researcher with a rather large pool of spontaneous self- statements. Since this method was generally incorporated immediately after a given experience, it would seem that self-talk recall would be most accurate following an event of short duration. It is likely that subjects would experience some difficulty when asked, for example, to recall and record everything they said to themselves during a lengthy sport event. Perhaps for this reason the production method has not yet been used in the athletic setting. In such situations, it might have to be combined with the recording method. Endorsement The endorsement method was the fourth category of self-talk assessment outlined by Kendall and Hollon (1981). This procedure 38 refers to a process in which subjects read through a list of self- statements to indicate how often, if at all, they emit each self- statement in a given situation. For example, Kendall, Williams, Pechacek, Graham, Shisslak and Herzoff (1979) designed a 20-item Self-Statement Inventory to examine the relationship between a patient's self-talk and subsequent ability to cope with a cardiac catheterization procedure. Kendall et al. collected a pool of rele- vant positive and negative self-statements from patients who had undergone this medical procedure. Forty college students coded each statement as to whether it would aid or hinder coping behavior during the cardiac catheterization. The only statements which were retained were those in which 100% consensual validation was attained. The final Self-Statement Inventory consisted of 10 positive (helpful) statements and 10 negative (hindering) statements. The final 20-item instrument was then administered to patients who had just undergone a cardiac catheterization. Subjects indicated on a 5-point scale the extent to which each statement was typical of their self—talk during the medical procedure. Results indicated a strong relationship between high frequencies of negative self-statements and inadequate coping behaviors. Hollon and Kendall (1980) also employed the endorsement method to develop an assessment tool, the Automatic Thoughts Questionnaire (ATQ-30), to examine the self-statements of depressed and nondepressed individuals. The investigators developed an initial pool of self- statements by asking a large number of college students to generate the type of self-talk they engaged in the last time they were 39 depressed. Next, these self-statements were given to psychometri- cally depressed and nondepressed individuals. They were asked to rate on a 5-point scale the frequency with which they emitted each of the statements within the previous week. By retaining the statements that significantly discriminated between the two criterion groups, the investigators ended up with a 30-item inventory. The ATQ-30 was sub- sequently cross-validated on a second group of psychometrically depressed and nondepressed individuals. A factor analysis with vari- max rotations revealed that four factors emerged with eigen values greater than 1. The investigators labeled the four factors as reflect— ing the following personal dispositions: Personal Maladjustment and Desire for Change; Negative Self-Concept and Negative Expectations; Low Self-Esteem; and, Giving Up/Helplessness. An analysis of the item responses revealed that both the depressed males and females responded with a higher frequency to each of the items on the ATQ-30, than did the nondepressed individuals. The advantage of using the endorsement procedure is that the assessment instrument does not have to be administered immediately after the situation under investigation. Since the assessment tool can be given in a neutral setting, it could be used to analyze self— talk relevant to a wide variety of problematic conditions. Further- more, the endorsement method may assist the subjects in self-talk recall because they are presented with self-statements, rather than having to generate them. A potential limitation of the endorse- ment method is that the self-statements which comprise the assessment 40 instrument may not be representative of the universe of self- statements from the selected population of subjects. Conclusions In view of the potential strengths and weaknesses of each of the four categories of assessment modes reviewed, the endorsement proce- dure appears to be the most practical method to employ for the current investigation which focused on the assessment of the automatic self- talk of tennis players. The use of the recording, sampling, and production procedures may be impractical to incorporate for the following reasons. Use of the videotape recording method would be too time consuming if incorporated in the competitive tennis setting. For certain team sports, an entire team could watch the tape simultaneously with each person watching only their performance, whereas with a sport such as tennis, only one person could be audio-video taped at a time and consequently, only one person could review the tape at a time. Furthermore, since the tape is generally played back immediately after the given situation, only one player per competitive meet would be taped and subsequently could review the tape directly after the con- test. Therefore, this method would be particularly inefficient with large numbers of tennis players. The audio-tape recording method enabled investigators to record the spontaneous overt self—talk of individuals as they performed a specified task. Even if tennis players could be hooked up with a cordless microphone, the players' attention could be distracted from 41 the task at hand, if they had to think outloud. Furthermore, as was mentioned earlier, the overt self-talk of an individual who is instructed to think out loud, may not accurately depict her/his ongoing covert self-talk. Consequently, neither of the recording methods seem practical for use in the competitive setting to assess tennis players' self-statements. The sampling method, in which subjects are cued at random time intervals to tune into their covert self-talk, would also be impracti- cal in the competitive tennis setting. It is not feasible to stop all play at random time intervals to have the athlete recall and record his/her self-talk. In fact, there is a tennis rule which specifies that play must continue without interruption. Even if play- ers could conceivably record their self-talk between games, the recording would not occur at random time intervals. The production method, as Kendall and Hollon (1981) define it, refers to the process in which subjects retrospectively recall and record their self-talk immediately after being exposed to the speci- fied situation. Since a tennis match often lasts an hour or more, it is likely that a player would experience some recall difficulty if asked to remember and record all match related self-talk. This problem could possibly be minimized if athletes were given specific game situations on which to focus their self-statement recall. The first step of the endorsement method actually employs a variation of the production procedure. That is, subjects are asked to think about the last time they were in a specified situation, and 42 then to recall and record everything they remember saying to them- selves at that point in time. The self-statements generated by this procedure then serve as the initial pool of statements from which the assessment tool is drawn. This procedure could be used to develOp an initial pool of competition-related self-statements of tennis players, provided that they were given specific prompts to aid in self-talk recall. The resultant endorsement“ instrument would then be developed by a process of consensual validation. It is difficult to determine validity empirically for each of the four major methods of self-statement assessment. However, the process of developing an assessment tool by use of the endorsement method insures that the final instrument will have content validity. It is possible that the use of retrospective reports employed by the endorsement. procedure may not be as valid, for example, as the use of on-the-spot recall as employed by the sampling method. The endorse- ment method, however, appears to be the most plausible mode of assessment to examine the competition-related self-talk of a tennis player, due to the lack of feasibility of the recording, production, and sampling methods. Furthermore, all studies completed in the field of clinical psychology which have incorporated retrospective reports as a means to gather self-statement data, report that subjects had little or no difficulty in recalling their situation specific self-statements (Cacioppo et al., 1979; Hollon 8 Kendall, 1980; Kendall et al., 1979; Schwartz 8 Gottman, 1976). In summary, very little clinically relevant research has been ‘ completed to date which has focused on the assessment of automatic 43 self-talk. Furthermore, only one sports-related study (Klinger et al., 1981) was designed specifically to examine the relationship between self-talk and competitive performance. Nevertheless, a grow- ing number of multifaceted intervention programs aimed at improving athletic performance have incorporated various methods of self- statement modification as an integral component (Gravel et al., 1980; Meyers 8 Schleser, 1980; Silva, 1980; and Suinn, 1980). Before the relationship between self-statement modification and subsequent behavior change can be assumed, assessment tools are needed first, to determine if self-talk can be measured, and second, if it can be measured, to assess an individual's internal dialogue from pre- to posttreatment for possible changes. Only then can the efficacy of various intervention programs aimed at self-statement modification be confirmed. CHAPTER III PROCEDURES The procedures for the development of the Self-Talk Question- naire (STQ-3), which are outlined in this chapter, were loosely patterned after the procedures that Hollon and Kendall (1979) used to construct the Automatic Thoughts Questionnaire (ATQ-30). Three major phases were employed to develop the STQ-3. The construction and subsequent administration of the STQ-1, the STQ-2, and the STQ-3 each constituted a phase. Thus, the development of each successive ques- tionnaire was dependent upon the results obtained from the adminis- tration of the previous instrument(s). Phase I: First Stage in the Development of STQ73 (STQ-1) The construction and administration of the STQ-1 represented the first phase in the development of the STQ-3. The purpose of this phase was to collect a pool of self-statements related to tennis com— petition. Thus, the STQ-l was designed to help tennis players recall and record their game related self-talk, to enable the investigator to develop a self-statement pool. Phase I thereby incorporated the following steps: (1) construction of the STQ-l; (2) construction of player rating scales; (3) selection of sample; (4) administration of 44 45 STQ-1; (5) collection of player skill and competitive pyschological effectiveness ratings; and (6) compilation of self-statements. Construction of STQ-1 Since the purpose of this phase was to collect a wide variety of tennis-related self-statements, four major game conditions were selected to aid players in self-talk recall within the STQ-1. These major game situations included: (1) Leading in a set; (2) Behind in a set; (3) Pleased with play; and (4) Displeased with play. In addi- tion, subdivisions of game situations were included under each of the four major game conditions to help players focus on specific aspects of their play. For example, under the playing condition of Pleased with play, players were asked to recall their self-talk when they were pleased with their groundstrokes, serves, volleys, and over- heads. It was assumed that the chosen game situations were general enough to allow players of varying ages, skill levels, and degree of psychological effectiveness to relate to them. Furthermore, it was also assumed that the four game conditions and respective subdivisions would stimulate the recall of a variety of self-statements, including positive, negative, on-task, and off-task items. Space was provided at the end of the STQ-1 so that subjects could add game conditions during which they frequently engaged in self-talk, along with their corresponding self-statements. Instructions for the STQ-1 were designed to aid players in their self-talk recall. Subjects were asked to recall a match they had played within the past week against an opponent of their sex and 46 comparable skill level. Next, they were asked to vividly visualize themselves competing under one of the four given game conditions, and then to write down what they remembered saying to themselves at that point. It was assumed that the retrieval of information stored in short and/or long-term memory would be aided by visualization techniques (Paivio, 1971) and situation specific retrieval cues (Ericsson 8 Simon, 1981). A copy of the STQ—1 is located in Appen- dix A. Selection and Construction of Player Rating Scales One of the two rating tools utilized in conjunction with the STQ-1 was the National Tennis Rating Program (NTRP). The NTRP is used nationwide by coaches and instructors to assess skill level and subsequently to assign players to apprOpriate classes and competitive events. The NTRP scale ranges from a low of 1.0 (beginning player) to a high of 7.0 (world class player). A copy of the NTRP can be found in Appendix B. The second rating tool utilized in conjunction with the STQ-1 was a Competitive Psychological Effectiveness (CPE) Scale. This scale was designed and incorporated to assess the degree of each player's competitive psychological effectiveness/ineffectiveness. In other words, the scale was designed to rate how steady, reliable, and con- sistent players were in the competitive setting, as compared to the practice setting. A 5-point scale was chosen as the most practical means to measure varying levels of competitive psychological 47 effectiveness. Of the odd numbered scales utilized by researchers, 3-, 5-, and 7-point scales have been the most common, with a 5-point scale used most often (Miller, 1973). The scale choices ranged from (5) Almost always playing to top of potential, to (1) Rarely playing to top of potential. An odd numbered scale was selected so that the middle number would depict those players who were approximately as psychologically effective as they were ineffective. The investigator intended to omit this middle group for certain statistical analyses, so that distinct criterion groups of psychologically effective and ineffective players could be formed. Once the 5-point scale was designed, the investigator sought the opinions of five area tennis coaches on the feasibility and clarity of the CPE Scale. Each individual asserted that the scale was easy to understand and would be easy to incorporate. A copy of this scale can be found in Appendix C. Selection of Sample Tennis players were required to meet three criteria in order to participate in Phase I of this study. First, the tennis players had to be at least 16 years old in order to minimize the confounding effects of an individual's cognitive deve10pmental processes. That is, ascording to Piagetian theory, the last stage of cognitive devel- Opment, formal operations, is attained approximately between the ages of 11 and 15 (Wadsworth, 1971). Even though an individual will con- tinue to learn beyond the acquisition of the formal operations stage, no major qualitative cognitive developmental changes should occur 48 (Brunk, 1977). Since it is possible that an individual's self-talk processes and subsequent behavior would be affected by qualitative cognitive developmental changes, players under 16 years of age were not asked to participate. The second requirement for subject selection in Phase I was a skill criterion. Subjects with a self-assigned NTRP skill rating of 3.0 or above were solicited to take the STQ-l. A tennis player with a NTRP 3.0 rating is described as one who "can place shots with moder— ate success; can sustain a rally of slow pace but is not comfortable with all strokes; lacks control when trying for power" (National Tennis Association, 1981). It was assumed that when subjects were asked to rate their own current level of play, some variation in the interpretation of the NTRP skill ratings would occur. However, Phase I did not require valid and reliable skill ratings across all subjects. The self-ratings were used primarily as a screening process to exclude players with a NTRP ratings lower than 3.0. The purpose of excluding these players was to minimize the number of technique- based, self-instructional statements often emitted by beginning tennis players. For example, an individual who is just learning a new motor skill may break the skill down into parts and instruct him/herself on each of these parts while executing that skill (Girodo, 1977). Hence, a beginning tennis player might instruct herself during a forehand drive to keep her eyes on the ball, to step across with her left foot, to hit in front, to follow through high, and to return to ready posi- tion. Girodo hypothesized that once the motor skills became more automatic, the number of technique-based self-instructional statements 49 would eventually be minimized or eliminated. He further proposed that if this did not occur, those self-statements could interfere with the smooth, rapid execution of the skill. The third criterion for participation required that the tennis players selected had to have played a competitive tennis match within one week prior to the STQ—1 administration. It was the intent of the investigator to administer the STQ-1 as soon after the players' com- petitive match as possible, to minimize the amount of information lost (Anderson, 1981). Furthermore, the investigator attempted to solicit players from a wide range of skill levels (NTRP scores from 3.0 to 7.0), ages (16 years and older), and geographical areas (Florida, Michigan, and Texas). An available sample of 29 males and 25 females was selected to participate in Phase I. It was assumed that these 54 individuals would generate a pool of more than 1,000 self-statements which would suffi- ciently represent the p0pulation of self-statements that tennis players 16 years and older engage in while competing. Both sexes were asked to participate in order to avoid biased results due to possible gender effects. Self—talk differences between males and females have been documented in the literature. For example, Weinberg, Gould, and Jackson (1979) examined the relationship between high and low levels of self-efficacy and corresponding motor performance outcomes. One of their findings indicated that overall, males engaged in sig- nificantly more positive self-talk and females engaged in significantly more negative self—talk while completing a motor task. 50 Administration of STQ-1 Tennis players and/or tennis instructors/coaches were either personally approached, contacted by telephone, or by mail. Individuals who were personally approached and who agreed to participate, were given on the spot step-by—step directions. Players were given the choice of taking the STQ-1 at that time or of taking it at a later date and bringing it back to a pre-arranged drOp-off point. Coaches/ instructors who were contacted, were given the option of allowing the investigator to administer the STQ-1 to their players, or, of adminis- tering it themselves at that time, or at a later date. The majority of coaches/instructors were contacted by telephone at which time the purpose of the investigation.was explained to them. If they agreed to allow their players to participate, an appointment was arranged with the coach/instructor, so that the investigator could personally explain the administration procedures for the STQ-1. Approximately half of the coaches/instructors chose to administer the STQ-1 themselves, at their convenience. The remaining coaches/instruc- tors permitted the investigator to administer the STQ-1 to their players. Those individuals who were contacted by mail received a brief explanation of the purpose of the study along with a request for their participation. The STQ—1, complete with directions, was enclosed along with a self-addressed, stamped envelope for their convenience. A total of 54 players, of whom 29 were males and 25 were females, participated in this initial phase of the investiga- tion. An overview of the characteristics of these participants 51 revealed wide ranges of both age and years of tennis experience, as Table 1 indicates. Ages ranged from 16 to 59 years, with a mean of 26.15. The mean number of years STQ-1 participants had been playing tennis was 7.64, with a range of 1 to 50 years. Collection of Player Skill (NTRP) and Competitive Psychological Effectiveness (CPE) Ratingg Although the need for player skill and competitive psychological effectiveness (CPE) ratings were not essential at this point, it was decided to incorporate these ratings as a part of the STQ—1 package to pilot test the feasibility of using these two scales for the STQ-2 and STQ-3. All participants were asked to rate their own skill level, according to the NTRP scale, and to rate their own level of competi— tive psychological effectiveness/ineffectiveness according to the CPE scale. A review of the self-ratings by these individuals indicated that all skill levels (NTRP of 3.0 through 7.0) were represented, with a mean of 4.24. Additionally, each level of psychological effectiveness/ineffectiveness was also represented with a mean CPE score being 3.54. Tables 2 and 3 display the frequency distribution of both the NTRP and CPE player self-ratings. Compilation of Self-Statements A total of 2,017 competition-related self-statements were col— lected. In addition to the items generated by the STQ-1, the inves- tigator also obtained situation appropriate statements from avail— able literature and informal interviews with tennis players. 52 Table 1. Phase I Frequency Distribution of Ages of Players and Years of Tennis Playing Experience Age Years of Experience Years f % f % 1—5 0 O 29 58 6—10 0 O 10 20 11-15 0 O 5 10 16-20 25 46 2 4 21-25 10 19 1 2 26-30 9 17 0 0 31-35 1 2 1 2 36-40 2 4 O 0 41-45 3 5 1 2 46-50 1 2 1 2 51-55 0 0 O 0 56-60 3 5 0 0 53 Table 2. Frequency Distribution of STQ-1 CPE Ratings CPE Rating Frequency Percentage 5 8 15 4 22 42 3 15 29 2 4 8 1 3 6 Table 3. Frequency Distribution of STQ-1 NTRP Ratings NTRP Ratings Frequency Percentage 3.0 9 17 3.5 12 23 4.0 9 17 4.5 9 17 5.0 5 10 5.5 4 8 6.0 1 2 6.5 2 4 7.0 1 . 2 54 Moreover, the investigator watched 26 tennis matches and unobtrusively recorded the overt self-talk of the players. Approximately 95% of the 2,107 statements collected were generated by the STQ-1. Phase II: Second Stage in the Development of the STQ73 (STQrZ) The purpose of Phase II was to reduce the size of the item pool and to use the remaining statements to construct a second self-talk questionnaire, the STQ-2. The STQ-2 was designed to reveal the fre- quency with which players emitted specific self-statements during a competitive event. Phase II then, incorporated the following steps: 1. elimination of redundant and inappropriate responses from the self-statement pool generated by the STQ-1 classification of statements into organized units to check for and eliminate additional redundant responses employment of raters to assess remaining statements so that the item pool could be further reduced by retaining only those items with 100% rater agreement construction of STQ-2 revision of CPE rating scale pilot testing of STQ-2 selection of sample administration of STQ-2 collection of player skill and competitive psychological effectiveness ratings The process of reducing the number of STQ-1 statements involved a variety of methodological procedures. A chart which depicts the 55 sequence of item elimination is presented in Table 4 to aid the reader in following the methods used to reduce the statement pool. Table 4. Sequence of STQ-1 Item Elimination for Construction of STQ-2 Total Number of Method of Item Elimination Statements Remaining Two judges independently eliminated redundant and nonsensical statements from the original item pool of 2017 statements (100% agreement required) 726 Two judges independently classified the remaining 726 statements into the categories created by Klinger et al. (1981) (87% agree- ment attained) Two judged independently eliminated redundant statements within each of the four game situa— tion categories (100% agreement required) 663 Five judges independently assigned the remain- ing 663 statements to effective/ineffective and skilled/unskilled categories (100% rater agree- ment across both the effective/ineffective 23d skilled/unskilled categories required for statement retention) 118 Elimination of Redundant and Inappropriate Responses The item pool of 2,017 competition-related self-statements was modified in several ways. First, the pool was screened for redundant, inappropriate, and unintelligible responses. Second, the specific subdivisions or focal points listed under each of the four major game situation headings (for example, describing self-talk when 56 executingaaforehand, or backhand, or overhead, etc., when pleased with play) were eliminated so that similar statements repeated under each of the focal points could be reduced to one representative statement. In addition, when necessary, the statements were deperson— alized and corrected for grammatical and spelling errors. The revised group of self-statements was then compared to the original set by an assistant to insure that the revised pool was representative of the original set of statements. Total agreement between the investigator and assistant was required before a state- ment was declared either redundant, inappr0priate, or unintelligible, and was consequently discarded. This process reduced the item pool to 726 statements. Classification of Statements The investigator subsequently reduced the item pool by categoriz- ing statements into organized units which allowed her to detect and consequently eliminate, additional redundant statements. On the basis of a review of related literature,the investigator chose the self- statement classification system incorporated by Klinger et al. (1981). The Klinger et al. investigation was the only available study in which the self-talk of athletes was categorized. Although these investiga- tors were scrutinizing the self—talk of basketball players, their categories of self-statement assessment were deemed apprOpriate for the classification of tennis players' self-talk. Klinger et al. employed the following five competition-related self-statement cate- gories: (1) completely unrelated, (2) irrelevant to good play, 57 (3) difficulty/ability, (4) self-exhortation/evaluation, and (5) problem solving/perceptual. These researchers reported that a 78% rater agreement was attained for the placement of statements into the aforementioned categories. The Klinger et al. classification system was thereby used in this investigation to categorize each of the remaining 726 STQ-1 statements. The investigator and an assistant independently classi- fied statements within each of the four given game situation cate- gories of Leading, Behind, Pleased, and Displeased. All statement classifications by the investigator and assistant were compared within each of the four categories, resulting in an overall rater agreement of 87%. Specifically, both the investigator and assistant achieved 86% agreement on classifying statements within the Leading category, 92% agreement within the Behind category, 79% for the Pleased category, and 91% for the Displeased category. An average of 45% of the statements across all four game situa- tion categories were classified as self-exhortations/evaluations (see Table 5). This was in contrast to the Klinger et al. results in which the overall distribution of statements was similarly spread among the last four content categories (16% irrelevant, 20% difficulty/ability, 21% self-exhortation/evaluation, and 27% problem-solving/perceptual). The contrast in findings could be accounted for by inherent differ- ences between the two Sports, methods of self-statement sampling, and/ or the game situations analyzed. Once the statements were categorized, the investigator and an assistant re-examined the item pool to pinpoint additional redundancies. 58 Table 5. Distribution of STQ-1 Self-Statement Classifications Game Situation Categories Self-Statement Categoriesa Leading Behind Pleased ‘ Displeased f 2 f 2 f 2 f z unrelated/ 24 17 35 20 11 6 32 15 Irrelevant Difficulty/ Ability 25 17 48 28 4o 21 57 27 salf‘EXh°rtati°n/ 71 49 71 41 82 43 101 47 Evaluation Pr°blem’s°1Ving/ 25 17 20 11 59 31 25 12 Perceptual -—-- -—- ———» -——- TOTAL 145 174 192 215 aCategories of self-statement classifications devised by Klinger et al. (1981). 59 Again, both investigators worked independently. A statement was sub- sequently discarded only if there was 100% agreement on its redundancy. Redundant statements were permitted among the four game situations, but not within a particular game situation category. This process enabled the investigator to reduce the number of statements from 726 to 663. Employment of Raters to Assess Remaining Statements It was the opinion of the investigator that the remaining pool of 663 items was still too large for an effective questionnaire, thus a decision was made to further reduce the number of items by retain— ing only those statements which might potentially discriminate between high and low skilled players, and discriminate between psychologically effective and ineffective players. Hence, five independent raters with backgrounds in both psychology and tennis were employed to classify each of the remaining statements. The 663 statements were given to the raters in four separate packets, with each packet con- taining statements emitted during one of the four game situations of Leading, Behind, Pleased,or Displeased. These individuals were then asked to examine every statement within each of the four packets and decide whether the statement was emitted (1) by a psychologically effective or ineffective player, and (2) by a skilled or unskilled player. If a rater was unable to decide which type of player (skilled/ unskilled and/or effective/ineffective) emitted a particular state— ment, the response "undecided" was recorded. Complete directions, statement packets, and coding forms can be found in Appendix D. 60 There was complete agreement among the raters in the assignment of 206 of the 663 statements to either the skilled or unskilled criterion groups. Additionally, raters were in 100% agreement in the placement of 321 of the 663 statements into either the psychologically effective or ineffective criterion groups. Since the purpose of the self-talk questionnaire was to distinguish between psychologically effective and ineffective players across a wide range of skill levels, the only statements retained for the STQ-2 were those which had sup- ported 100% rater agreement across both the skill/unskilled and psychologically effective/ineffective criterion groups. This process reduced the item pool from 663 to 118 statements. As a result, the four major game conditions of Leading, Behind, Pleased, and Displeased within the STQ-2, contained 25, 34, 27, and 32 statements, respectively. Construction STQéZ The STQ-2, which is located in Appendix E, consisted of 118 state- ments subdivided under the major situation headings. All items were randomly ordered within each of the four major categories. Addition- ally, the presentation order of the four game situations was rotated in four ways among the questionnaire packets, to minimize the effect, if any, of order on the results. That is, intuitively it seemed plausible that a response bias might occur if the leading category, for example, was always given first and the displeased category was always given last. A 5-point rating scale was incorporated in the STQ-2 so that players could indicate the frequency with which they emitted a 61 particular statement to themselves, either covertly or overtly, while competing under one of the four major game conditions. The rating scale, ranged from 1 = NOT AT ALL, to 5 a ALMOST ALWAYS. The use of a 5-point frequency scale was selected for a number of reasons. The recommended range for such scales is from three to nine units, with a 5-point scale used most often (Miller, 1973). An odd number of choices are preferred so that the middle number can serve as an aver- age rating (Verducci, 1980). A number of investigators who have assessed the frequency with which an individual emitted various self- statements have effectively utilized a 5-point rating scale (Hollon 8 Kendall, 1980; Kendall et al., 1979; Schwartz 8 Gottman, 1976). Con- sequently, a 5-point frequency rating scale was selected for use in this study. Although the 5-point rating scale used for this investigation was patterned after that of Hollon and Kendall's (1980) ATQ-30 rating scale, descriptive statements for three of the five rating selections were altered. First, the descriptive phrases for the second and third ratings of "sometimes" and "moderately often" were changed to ' respectively. It was the opinion of "infrequently" and "sometimes,' this investigator that such descriptors would serve as better dis- criminators. In addition, the fifth rating, "all the time" was deemed inappropriate since presumably individuals would seldom repeat the same self-statement to themselves 100% of the time, with no interference from other self-statements. Therefore, the phrase "almost always" was chosen to depict a high rate of self-talk fre- quency in the current investigation. 62 Revision of CPE Rating Scale Minor alterations of instructions and descriptive components within the CPE scale were mandated due to player responses to the scale given in conjunction with the STQ-1. Post-questionnaire inter- views indicated that a number of players had confused the CPE and skill ratings with each other. Hence, the instructions and scale descriptions were modified to clearly distinguish between the two types of ratings. A copy of the revised CPE scale incorporated in the STQ-2 can be found in Appendix F. Pilot Testing of STQ+2 The STQ-2 was pilot tested on nine tennis players of varying ages and educational backgrounds. Each player was given the STQ-2 with no verbal instructions other than to "Follow the written direc- tions." Upon completion of the task, players were asked about the clarity of instructions, the ease or difficulty of following the directions, and their understanding of both the psychological effec- tiveness, and skill rating scales. All subjects indicated that the directions were clear and easy to follow. Furthermore, they all gave accurate interpretations of the two rating scales. Based on these positive findings, the investigator assumed that teachers/coaches would be qualified to administer the STQ—2. Selection of Sample Following the pilot test for the STQ-2, the investigator attempted to select a relatively heterogeneous sample of players to 63 participate in this next phase. Subjects for Phase II were chosen from teams located in lower Michigan. More specifically, individuals who did not participate in Phase I, and met the following four cri- teria were eligible to participate in Phase II. First, players had to be at least 16 years old. Second, they had to be a member of a club, college, or high school sponsored tennis team or class. Third, they had to receive a NTRP rating of 3.0 or higher from their coaches/ teachers. Fourth, players must have played a singles competitive tennis match within one week prior to the STQ-2 administration. The investigator selected as many subjects as possible for this second phase to ensure that their number minimally equaled the total number of self-statements, plus two (Klecka, 1980). Such a rule of thumb minimized problems for data analysis and interpretation. Administration of STQ-2 Coaches/instructors were contacted by phone at which time the purpose of the study was explained to them. If they agreed to par- ticipate, an appointment was arranged so that the investigator could explain the administrative procedures to the coach/instructor. In the majority of cases, the coach/instructor chose to personally administer the questionnaire to their players. The STQ-2 was distributed to 407 players from lower Michigan who met the eligibility requirements for participation. The use of extensive follow-up measures, including multiple telephone calls, personal visits, and written reminders, resulted in the return of 179 questionnaires (42% return rate). Of the 179 questionnaires, 64 17 were discarded because of multiple missing responses. An addi- tional 11 questionnaires were also omitted from the data analyses because the players' ages were under 16. Thus, a total of 151 ques- tionnaires were used for data analyses. A review of the demographic information provided by the remain- ing 151 players revealed that a disproportionate number of males (45) and females (106) had completed the STQ-2. Since the original Phase II sample of 407 players was comprised of approximately equal numbers of males and females, the disproportionate sample result appeared to be a function of more females taking the time to complete and hand in the STQ-2 than males. Each of the 151 individuals who participated in this phase was a member of one of 12 tennis clubs located in six lower Michigan cities. Phase II players included junior players (23%), traveling team mem- bers (50%), tennis professionals (9%), high school players (6%), and adult club members (13%). Ages of these players were widely dispersed, ranging from 16 to 55 years, with a mean age of 31.74. The playing experience of these individuals was also widely varied, which ranged from 1 to 41 years, with a mean of 7.48 years. A frequency distribu- tion which displays the ages, and years of playing experience of the Phase II participants is given in Table 6. Collection of Players Skill and Competitive Psychological Effectiveness Ratings In addition to rating the 118 self-statements, players were also asked to assign themselves a skill rating (NTRP) and a competitive 65 Table 6. Phase II Frequency Distribution of Ages of Players and Years of Tennis Playing Experience Age Years of Experience Hours f % f % 1-5 0 0 81 54 6-10 0 O 41 27 11-15 0 0 12 8 16-20 43 28 9 6 21-25 6 4 4 3 26-30 18 12 3 2 31-35 20 13 36-40 28 19 41-45 14 9 46-50 15 10 51-55 7 5 66 psychological effectiveness (CPE) rating. Furthermore, coaches/ instructors were asked to rate each of their players according to both the NTRP and CPE Scales. (Complete instructions for the coaches/ instructors, along with a copy of the NTRP and CPE Scales, can be located in Appendixes B and C, respectively.) Coaches/instructors were asked to rate each of their players twice on the CPE Scale. These ratings were separated by at least one day. This procedure enabled the investigator Ix) establish the reliability of this newly created scale. The correlation coefficient computed between the first and second CPE ratings by the coaches/instructors yielded a value of .9009 UR < .001), thereby suggesting a relatively high degree of scale reliability. In addition, a correlational analysis was employed to examine the relationship between the CPE ratings by the coaches/instructors and those by the players. A moderate relationship, r = .6417 (p.< .001), was reported. Based in part on this moderate correla- tional findings, and in part on logical reasoning, the decision was made to use the first CPE rating by the coaches/instructors rather than those by the players for all analytical procedures. It was assumed that the ratings by the coaches/instructors served as a more consistent measure of player effectiveness than did the individual player self-ratings. That is, the ratings by one individual of 20 players, for example, might be more consistent than 20 self-ratings by 20 players, with each rating perhaps influenced by unique interpreta- tions. Furthermore, it was felt that the coaches/instructors' ratings Inight be less biased than the players' self-ratings. 67 The relationship between the NTRP ratings by the coaches/ instructors and by the players was also analyzed. The results of a correlational analysis revealed a strong correlation coefficient of .9382 (p < .001). Since the NTRP scale is commonly used nation- wide by instructors and coaches to place players in appropriate skill level classes and competitive events, and since the correla- tion beween the coaches/instructors and players' self-ratings was high, it was concluded that the classification of skill on the basis of the NTRP scale was adequately accurate and reliable for use in this investigation. An examination of the frequency distribution of both coach- assigned CPE and NTRP ratings revealed that all five CPE levels and all nine NTRP levels were represented by these Phase III participants (see Tables7'and 8). Mean CPE and NTRP scores were 3.35 and 4.29, respectively. Table 7. Frequency Distribution of STQ-2 Coach Assigned CPE Ratings CPE Rating Frequency Percent 5 10 7 4 62 41 3 56 37 2 17 11 68 Table 8. Frequency Distribution of STQ-2 Coach Assigned NTRP Ratings NTRP Rating Frequency Percent 3.0 26 17 3.5 37 25 4.0 24 16 4.5 15 10 5.0 13 9 5.5 18 12 6.0 ' 6.5 7.0 Phase 111: Third Stage in the Development of the STQ-3 The purpose of the third and final stage in the development of the STQ-3 was to statistically analyze the STQ-2 responses, and to select only those statements which effectively discriminated between the psychologically effective and ineffective criterion groups to comprise the STQ-3. Hence, Phase III included the development of the STQ-3: (1) examination (2) analysis of STQ-2 responses and subsequent pool, (3) construction of STQ-3, (4) selection tration of STQ-3, and (6) collection of player ness ratings. Examination of Order Effect the following steps in of order effect, reduction of item of sample, (5) adminis— skill and effective- The investigator computed a multiple analysis of variance (MANOVA) as a precautionary measure to insure that the four presentation orders 69 of the game situation categories (Leading, Behind, Pleased, Dis- pleased) had not affected player responses. A summary of the MANOVA results is shown in Table 9. Table 9. Results of Wilks' Lambda to Assess the Effect of Category Order on Player Responses Game Wilks' Approximate df Significance Situation Lambda F of F Leading .81233 1.11817 25,121 .33340 Behind .75789 1.06170 34,113 .39497 Pleased .80387 1.06628 27,118 .39109 Displeased .80556 .84480 32,112 .70213 As can be seen from Table 9, levels of significance for the Wilks' lambda test ranged from .33 to .70, thereby indicating that the various presentation orders within the STQ-2 did not significantly (p :>.05) affect the players' responses. In view of these satisfactory results, a detailed, multifaceted analysis of the STQ-2 player responses was initiated. Analysis of STQ-2 Reaponses The investigator performed a variety of statistical analyses to determine which items would comprise the final questionnaire, the STQ-3. The results of discriminant, factor, correlational, and reliability analyses were used to select a total of 51 statements. 70 An overview of the statistical analyses employed and the sequential elimination of 67 STQ-2 statements is outlined in Table 10. The major purpose of this investigation was to construct an instrument which differentiated between the competition-related self- talk of psychologically effective and ineffective players. Conse- quently, two discriminant function analyses were employed to deter- mine which of the 118 STQ-2 items differentiated between various com- binations of CPE criterion groups. The alpha level for these analyses was set at .05. Since this was an exploratory investigation, a rather liberal cut off point for alpha was used, to retain as many poten- tially discriminating items as possible. The investigator formed criterion groups for the first discrim- inant analysis by assigning players with a CPE rating of 4 or 5 to the psychologically effective criterion group (p_= 68) and those with a CPE score of 1 or 2 to the psychologically ineffective criterion group (p_= 23). The independent variables in this case were the CPE criterion groups, whereas the dependent variables were the player responses to the STQ-2 statements. The null hypotheses of testing for identical mean vectors between the two criterion groups was assessed by the Wilks' lambda test. Since Wilks' lambda represents the percent of unexplained variance of the total variability, a low value was desired. The results revealed that the Wilks' lambda values for the 118 statements ranged from a low of .73240 (p < .00001) for for statement B34 ("I'm no good"), to a high of .99999 (p < .9794) for statement P27 ("Ah, perfect height"). Refer to Appendix H for the table which gives a complete listing of the Wilks' lambda values for 71 Table 10. Sequence of STQ-2 Statistical Analyses and Item Elimination for Construction of STQ-3 Total Statements d El Sequence of Analyses an Item imination Remaining Discriminant Analysis A combined total of 45 statements significantly discriminated (p < .05) between CPE criterion groups of l and 2 versus 4 and 5, and/or 1 and 2 versus 5 118 Correlational Analysis A total of 56 items significantly correlated (p < .05) with the CPE scores Elimination of Items A combined total of 60 items did not signifi- cantly discriminate between, and/or corre- lated with the CPE scores (p < .05); the remaining‘58 statements were retained 58 Factor Analysis Two factors emerged; 14 items represented the Appropriate Factor, and 44 items represented the Inappropriate Factor Reliability Analyses Cronbach's alpha was computed for the pool of 58 items,as well as for the pool of 56 statements (derived from the correlational analysis) and for the set of 34 items (derived from the discriminant analysis of the CPE criterion groups of 1 and 2 versus 4 and 5) Elimination of Items A comparison of the Factor I alpha coeffi- cients indicated that six items could be deleted from the Inappropriate Factor without substantially decreasing alpha 52 Final elimination of one potentially confusing item from within the Inappropriate factor 51 72 all 118 statements. An examination of the univariate F-ratios (df = 1,89) indicated that 34 of the 118 statements significantly dis- criminated (p < .05) between the psychologically effective and ineffective criterion groups. Since this was an exploratory investigation, self-talk patterns for the various levels of player CPE were not known. Hence, it was conceivable that players who were ranked at the extreme end of the scale (for example, CPE rating of 5) might respond somewhat differ- ently to the STQ-2 items than players who were assigned a CPE rating of 4, even though both player types had been previously defined as effective players. Thus, to provide supplementary information, a second discriminant analysis was computed based on the responses of 33 subjects, 10 who were assigned a CPE rating of 5, and 23 who were given a CPE rating of 1 or 2. The Wilks' lambda values for the 118 statements in this analysis ranged from a low of .60652 (p < .0001) for statement L7 ("1 need one more point to win, and I miss an easy shot. Darn, I choked."), to a high of .99997 (p_< .9745) for state- ment B23 ("Be aggressive, not cautious"). Results indicated that 35 of the 118 statements significantly discriminated (p.< .05) between the responses of players with a CPE rating of 5 and the responses from players assigned a CPE rating of 1 or 2. Eleven of these 35 state- ments, however, did not significantly discriminate between the origi- nal criterion groups based on CPE player ratings of 4 or 5 and 1 or 2. Hence, a combined total of 45 statements significantly discriminated between one or both of the effective criterion groups and the ine vi‘ ca: vi of SI CE 11C Si U. .’_7 73 ineffective criterion group. Complete listings of the 45 statements with their respective group univariate F values and levels of signifi- cance for both series of discriminant analyses are given in Table 11. It would have been equally beneficial to compute a parallel analysis on those players with a CPE rating of 1, compared to those with a CPE rating of 4 or 5. However, the disproportionate number of subjects in each of these criterion groups (six subjects with a CPE rating of 1 versus 68 subjects with a CPE rating of 4 or 5) made statistical group comparisons inappropriate. Although the results of the discriminant analysis provided use- ful information,the accuracy of those results was interpreted with caution because the requirement for a necessary and sufficient n was not met. Klecka (1980) recommended that there be at least two more subjects than the number of discriminating variables. The STQ—2 con- sisted of 118 statements (discriminating variables). Thus, a minimum of 120 subjects would have satisfied the requirement for an adequate sample size. Although 151 subjects completed the STQ-2, that number was reduced to 95 and 33 for the two discriminant analyses respectively, due to the omission of responses from players with a CPE rating of 3. A follow-up analysis was employed as a safeguard in the event that the discriminant analyses may have yielded results which were too liberal or conservative due to the insufficient n. Even if an adequate sample size would have been employed, a follow-up test would have been deemed appropriate, since a discriminant analysis will not choose items which are linear combinations of each other. 74 Table 11. STQ-2 Statements Which Significantly Discriminated* Between the CPE Criterion Groups . b 1 Itema “Criterion Groups ::;::a F Significance 1 or 2: 1 or 2:. 5 4 or 5' L3 X .86439 4.71 .0381 L6 X .95526 4.17 .0441 L7 X .60652 19.46 .0001 X .79691 22.68 .0000 L10 X .88442 3.92 .0569 X .92230 7.50 .0075 L11 X .88643 3.84 .0593 L15 X .83809 5.80 .0224 L16 X .75845 9.55 .0043 X .92509 7.21 .0087 L17 X .87401 4.33 .0462 X .94779 4.90 .0294 L19 X .95987 3.72 .0569 L20 X .95810 3.89 .0516 L21 X .64659 16.40 .0003 X .87771 12.40 .0007 L22 X .88554 3.88 .0582 Bl X .87699 4.21 .0491 B2 X .75560 9.70 .0040 X .92033 7.71 .0067 B4 X .88672 3.83 .0596 BS X .63557 17.20 .0003 X .84360 16.50 .0001 B11 X .92140 7.59 .0071 313 X .85557 5.06 .0319 B14 X .86365 4.74 .0375 B15 X .82130 6.53 .0159 X .89038 10.96 .0013 B17 X .83184 6.06 .0198 X .89771 10.14 .0020 319 X .85042 5.28 .0028 X .94673 5.01 .0277 320 X .85724 5.00 .0330 X .93533 6.15 .0150 B22 X .84953 5.31 .0283 X .82562 18.80 .0000 824 X .89406 10.55 .0016 326 X .93393 6.30 .0139 (table continues) P Dd PPDUDU hUnIawluI Table 11. Continued. 75 Criterion Groups Wilks' Item 1 or 2: 1 or 2: Lambda F Significance 5 4 or 5 328 X .84334 5.57 .0249 X .75369 29.09 .0000 333 X .61810 18.54 .0002 X .81312 20.46 .0000 334 X .75671 9.65 .0041 X .73240 32.52 .0000 P4 X .82858 6.21 .0185 X .93548 6.14 .0151 P5 X .81071 7.01 .0128 X .93610 6.08 .0156 P7 X .88195 4.02 .0542 P14 X .80124 7.44 .0105 P15 X .77239 8.84 .0058 P18 X .85972 4.90 .0347 D5 X .93931 5.75 .0186 D8 X .81857 6.65 .0151 X .90006 9.88 .0023 D11 X .74049 10.51 .0029 X .88183 11.93 .0008 D12 X .93880 5.80 .0181 D15 X .86773 4.57 .0407 D17 X .94918 4.77 .0317 D20 X .74554 10.24 .0032 X .90344 9.51 .0027 D24 X .75086 9.95 .0036 X .92417 7.30 .0082 D28 X .65361 15.90 .0004 X .84709 16.07 .0001 D31 X .96019 3.69 .0579 3The letter which precedes the STQ-2 item number indicates the game situation category to which the statement belongs: a set; 3 B behind in a set; P 8 pleased with play; D = displeased with play. b 2 versus 4 or 5. * p.:_.05. L = leading in Criterion groups based on CPE ratings of 1 or 2 versus 5, and, lor 76 Thus, if two or more statements, which were linear combinations of each other, significantly discriminated between the CPE criterion groups, only the best discriminator would have been selected by the discriminant analysis process. It was probable, therefore, that a number of potentially useful discriminating statements may have been unwantedly overlooked. To compensate for the aforementioned potential problems, the investigator subjected the data to a correlational analysis as a follow-up and as a supplemental piece of information to the discrimi- nant analyses. Spearman's rho, a nonparametric correlational tech- nique which utilizes data in the form of ranks, was employed to examine the relationship of each player's CPE rating to their respec- tive STQ-2 item responses. The ranks of player CPE scores were com- pared to ranks of player responses to each STQ-2 item. There were two major advantages of using Spearman's rho as a follow-up measure. First, rho was calculated for each STQ-2 item, thus leaving the deci- sion of whether or not to retain or delete items up to the investi- gator. Second, this analysis also permitted the scores of all 151 subjects to be utilized, rather than only those in the extreme groups. The resulting coefficients, presented in Table 12, indicated that 56 of the 118 statements correlated significantly (p < .05) with the coach/instructor assigned CPE rating. The relationship between the degree of psychological effectiveness and statement responses ranged from r8 = .0032 (p < .485) for statement D22 ("Find the timing spot") to rS = -.3766 (p < .001) for statement L7 ("I need one more point to win, and I miss an easy shot. Darn, I choked"). Table 12. STQ-2 Statements which Significantly Correlated* with the Five Levels of CPE Itema rho Significance Item rho Significance L2 -.1320 .054 322 —.3182 .001 L3 .1895 .010 324 -.1622 .024 L4 -.1555 .029 B26 -.1861 .012 L5 .1606 .025 328 -.3127 .001 L6 -.1919 .010 329 .1380 .046 L7 -.3766 .001 333 -.3601 .001 L10 -.2837 .001 334 —.3500 .001 L11 -.1487 .035 P2 -.1360 .048 L15 -.1324 .053 P3 .1342 .040 L16 -.2849 .001 P4 -.2448 .002 L17 -.2281 .003 P5 -.2184 .004 L19 .2336 .003 P10 .1611 .025 L20 .2271 .003 P14 -.1479 .035 L21 -.3135 .001 P15 -.1755 .016 L22 .2528 .001 P18 -.1373 .047 31 -.2144 .005 P24 .1407 .043 32 -.2643 .001 D4 -.1955 .009 33 -.1543 .030 35 -.1494 .034 34 -.1341 .051 D8 -.2702 .001 35 -.3109 .001 311 -.2515 .001 311 .2701 .001 312 -.2562 .001 312 .1803 .014 315 -.1500 .034 313 -.1520 .032 317 -.1905 .010 314 -.1577 .027 320 -.2586 .001 315 -.2780 .001 324 -.2837 .001 317 -.2843 .001 325 -.1406 .043 319 -.2123 .005 327 .1499 .034 320 -.1867 .012 328 -.3426 .001 8The letter which precedes the STQ-2 item number indicates the game situation category to which the statement belongs: *R<. U’Utfit“ leading in a set behind in a set pleased with play displeased with play 78 Thirty-three of the 56 statements which significantly corre- lated (p_< .05) with the coach/instructor's CPE rating, also signifi- cantly discriminated (p < .05) between the CPE groups of 4 and 5 versus 1 and 2. Similarly, 34 of those statements also significantly dis- criminated (p < .05) between CPE groups of 5 versus 1 and 2. Overall, a total of 58 statements significantly discriminated and/or corre- lated with the assigned CPE ratings. The 60 statements which neither significantly discriminated between, and/or correlated with, the CPE criterion groups were subsequently eliminated, since the ultimate goal of the STQ-3 was to distinguish between the self—talk emitted by psy- chologically effective and ineffective tennis players. Table 13 gives a complete listing of all those statements which were omitted, within each of the four game situation categories. The responses associated with the remaining 58 statements were further scrutinized by means of a factor analysis. The exploratory analysis was employed to determine if the statements could be grouped together by specific underlying constructs. A principal components factor analysis with iterations was used. Since the factor structure was clearly identified and meaningful with this method, no rotation of the factor matrix was deemed necessary for this phase of the investigation. An item was considered to belong to a factor if its loading was greater than or equal to .30. The investigator chose .30 to serve as a liberal cut-off point, since the major intent of using the factor analysis was to classify the remaining statements, rather than to delete them. Furthermore, only those factors which had eigenvalues 79 Table 13. Omitted STQ-2 Items Based on Nonsignificant* Results of Discriminant and Correlational Analyses Itema Statement L1 I know I can beat her/him. L8 Play as if the score were close. L9 Don't go crazy on shots, play it safe; play percentage tennis. L12 I have to play well. L13 Keep it up. Don't let down. You've got the momentum. L14 Play her/his weakness. L18 His/her backhand/forehand is weak. L23 I have to keep coming to the net to attack. L24 I wonder if I'm going to win this match. L25 Don't beat yourself. Make him/her win the point. Don't give it away. 36 Relax. Hit out. 37 Let's look at what's happening here. 38 Why did I hit that shot? 39 Concentrate! Get back in the game. That game is not over. 310 Stay with it. Play her/his faults. 316 I need to hold serve. 318 Try something new. Find a weakness; change the game. Make something happen. 321 Don't rush the passing shot. 323 Be aggressive, not cautious. 325 It's only a few points, don't let down. 327 Press a little bit, but don't make mistakes. 330 Stay cool. Let him/her make the mistakes. 331 Keep the ball deep. Stay in the point. 332 Stay in the point as long as possible. Push. Pl Make him/her work for it; make him/her move. P6 Get on top of the ball; hit on the rise. P8 Stay alert. Keep feet moving. P9 No change. Keep it up. P11 Watch his/her toes. P12 I don't know what I'm doing right, but I h0pe it lasts. P13 Good powerful shot; good speed and depth, nice and solid. P16 Wider; play the angles. P17 My forehand is good. I'm hitting nice and deep with a lot of top spin. P19 Look where she/he is standing. P20 I'm on and off. P21 Don't rush the passing shot. (table continues) 80 Table 13. Continued a __ Item Statement P22 Attack on his/her second serve. P23 P25 P26 P27 D1 D2 D3 D6 D7 D9 D10 D13 D14 D16 D18 D19 D21 D22 D23 D26 D29 D30 D32 Okay. Mix it up. She/he can't handle my hard forehands/backhands. That's more like it. Now a little more stuff on the first serve. Ah, perfect height. Stay solid/low. She/he must have loved that . . . burning me. Take your time. Slow the game down. Don't rush it. Lousy. What is my opponent thinking. Adjust the toss. Hit short, deep, down the line, cross-court, to the corner; etc . Shorten your backswing. You're overplaying the ball. Hit/aim higher; lift more; place the ball up; hit up. Racket in full follow through position. Watch the angle of the racket head. Next time. Forget that one. Hit with confidence. Be deliberate. You're playing too loose. Find the timing spot. Just one point at a time. Step into the volley. Stay back. Watch the seams on the ball; watch the spin. Move faster; move your feet; stay on your toes. a . The letter which precedes the STQ-2 item number indicates the game situation category to which the statement belongs. fp < .05. leading in a set behind in a set pleased with play displeased with play :3va llll 81 of greater than 1.00 were examined. A total of 14 factors met this criterion and were subsequently scrutinized to see if the statements within each of these factors were logically related to an underlying construct. Factors III through XIV exhibited nonsensical constructs. Hence, those 12 factors were eliminated from future analyses. Factors I and II, however, displayed very clear constructs, and in combination, explained 36.2% of the total variance. Consequently, the STQ-2 was judged to contain two major underlying constructs. Factor I accounted for 26.7% of the variance,with an eigenvalue of 14.94997, and Factor II accounted for 9.5% of the variance with an eigenvalue of 5.31194. Factor 1, labeled the Inappropriate Factor, was characterized by low self-efficacy statements, including self-critical and off-task ver- balizations. Factor II, labeled the Appropriate Factor, was character- ized by on-task, constructive self-directions, as well as words of encouragement and praise. A complete listing of statements under both factors, including the factor loadings for each statement is given in Table 14. The discriminant, correlational, and factor analyses were used to examine and subsequently support the extent of STQ-2 validity up to this point. In summary, the results of the discriminant analyses indicated which items differentiated between the CPE criterion groups; the use of Spearman's rho revealed the correlations between the levels of CPE and item responses; and finally, the results of the factor analysis disclosed two underlying constructs among the patterns of player responses. 82 Table 14. Factor Loadings of Remaining 58 STQ-2 Items on Factors I and II Factor Loadings fifiggér Self-Statement I II Factor I8 D11 Typical! Here I go again. Calm down. .70 .05 L7 I need one more point to win, and I miss an easy shot. Darn, I choked. .68 .03 L17 It's game/set/match point. Now don't blow it like you usually do. .68 -.10 326 Oh brother! Why am I playing so bad? I know that I can play better. .67 .22 D15 I hope I don't hit this out/into the net. .66 .00 B2 Darn, I blew it again. .66 .04 P4 Finally, you're doing something right; It's about time. .65 -.05 333 I'm choking; I'm blowing it. .64 -.13 P5 I hope that I can keep this up. .64 -.02 D28 This is not my day. .64 .02 L4 I can't believe I didn't mess it up. .63 -.03 D20 I feel foolish. Why don't you con- centrate? .62 .16 P15 It feels good. Why can't I do this all the time? .61 .11 P2 Thank goodness, once in awhile I can do it right. .61 .00 L6 I wonder how long my luck will hold out. .61 -.20 L21 I can't believe I'm really winning. .60 -.17 D8 What a screw up. There's no reason for that! I made a careless mistake. .59 .23 35 What is my problem? .59 .20 320 Oh darn! Please let me win just this once. .59 -.10 D12 My opponent must think I'm a real loser. .59 -.15 P14 I'm surprised I made that shot. .59 -.08 322 I hope no one I know is watching me. .58 -.12 317 I'll be glad when this match is over. .58 -.13 324 Maybe I'll get lucky and get it together. .57 -.09 P18 Alright! A wide open court. Don't - blow it. .56 .13 313 People are watching me. I wonder what they're thinking? .56 .12 328 I suppose you are going to give her/him the next point too? .55 -.09 (table continues) 83 Table 14. (Continued) Factor Loadings NEESer Self-Statement I II L15 I hope that I can win this next point/game. .54 .13 334 I'm no good. .54 -.31 L10 I'm ahead now, but it probably won't last. .54 -.24 L16 Don't mess up now. .54 .12 319 He/she is a lot better than me. .53 -.09 P7 It feels good but scary. I don't know if I can keep it up. .53 -.07 315 I never win. .52 -.16 33 There is no wayIIwill catch up. .52 -.27 31 Stupid; Fool; Idiot; Dummy; etc. .51 .08 314 The harder I try, the more I mess up. .51 .21 D24 I hope I don't double fault. .50 -.15 D25 Oh brother, right to his/her forehand. I'm going to get killed. .50 -.07 D4 If I could just have a good serve, I could win more. .49 -.03 L11 I hope I don't lose. .49 -.13 L2 I hit a lucky shot. Things are going - my way for a change. .48 -.05 D17 Oh no, it's a short shot. .40 -.08 34 I need a break. .36 .18 Factor IIb L19 Play your game. .12 .67 P10 Good, put the pressure on her/him. -.04 .63 329 Get going; Let's go; Go for it; Go to it. -.03 .62 L20 Stay loose, play each point. .07 .62 311 I have to come back and play wellg to win the match. .26 .62 L5 Concentrate on execution. .02 .59 312 Not too many outright winners, but when opportunity knocks, be ready. .18 .58 L22 Close her/him out. .05 .56 L3 He/she is tough, but I'm better. .15 .51 P3 Early preparation. -.01 .50 D27 OK, I made a mistake. Forget it. Concentrate on the next point. .02 .47 (table continues) 84 Table 14. (Continued) Factor Loadings NSIm Self-Statement I II D31 3/4 pace with depth. I'm hitting too hard. .22 .46 P24 Watching the ball all the way pays off. .07 .44 D5 You're catching it late. .21 .30 8I = Inappropriate Factor bII = Appropriate Factor Note. Percent variance for Factors I and II equaled 26.7 and 9.5, respectively. 85 As assessment of validity should be accompanied by an appraisal of reliability. A variety of methods were available to test the reliability of the STQ-2, including split-halves, alternate form, test-retest, and Cronbach's alpha. Carmines and Zeller (1979) sug- gested that there are several problems associated with the use of the split-halves, alternate form, and test-retest methods of assessing instrument reliability. The potential weakness of using the split- halves method in this study was that the value of the reliability coefficients could vary, depending on how the items were split up. The alternative form method was impractical since it required the construction of a parallel, alternative instrument. Furthermore, the disadvantage of the test-retest method was a possible confounding effect of the subjects' responses on the retest, from the first admin- istration of the instrument. For example, on one hand, a low test- retest correlation could have been a function of reactivity. That is, the very process of taking the STQ-2 for the first time might have raised the player's awareness of his/her self-talk patterns and thereby would have influenced his/her response on the retest. On the other hand, a high test—retest correlation might have reflected the player's response recall from the first to the second STQ-2 adminis- tration. Nunnally (1964) theorized that the subject's memory could inflate the test-retest reliability coefficient up to one month after the original test administration. If the administration of the retest were delayed for longer than one month, this investigator would have been faced with the possibility of confounded results due to subject 86 maturation, which, in turn, would likely have underestimated the projected reliability of STQ-2. In addition to the statistical difficulties associated with the employment of the test-retest procedure, more importantly, the type of self-talk measured may have also limited the usefulness of a test- retest method. The cognitive phenomenon under investigation could have fluctuated somewhat from situation to situation, thereby render- ing the employment of a retest impractical. That is, a multitude of external factors, such as opponent's skill, match importance, weather, court, and crowd conditions, as well as internal factors, such as fatigue, illness, injuries, personal problems, motivation, anxiety, and pressure could have separately or in combination, contributed to self- talk differences on a retest. Although a certain pattern of cross- situational cognitive consistency may have been likely, the STQ-2 retest responses could have varied enough to result in an underestimate of the instrument's level of reliability. In view of the potential problems associated with the various methods of reliability assessment, this investigator selected Cron- bach's alpha to examine the extent of the reliability for the remaining 58 statements. Cronbach's alpha, an estimate of the correlation between the original instrument and an alternative hypothetical instrument with the same number of items, incorporates both the Kuder- Richardon 20 and the Spearman-Brown prophecy formulas (Carmines 8 Zeller, 1979). Cronbach's alpha is considered a conservative measure of internal consistency, with the actual scale reliability never 87 falling below the alpha coefficient (Carmines 8 Zeller, 1979). Table 15 gives Cronbach's alpha for the set of 58 statements, as well as for each of the two factors analyzed separately. Table 15. Cronbach's Alpha Coefficients* for the Remaining 58 STQ-2 Itemsa Factors I & II I II Alpha No.c Alpha No. Alpha No. .94 58 .96 44 .85 14 aItems which significantly discriminated (p < .05) between and/or correlated with the CPE criterion groups. bFactor I represents the Inappropriate items; Factor II represents the Appropriate items. cNumber of items analyzed. *2 < .0001. As it can be seen from Table 15, the alpha coefficient for the remaining 58 statements was equal to .94 (p <1 .0001). Carmines and Zeller (1979) suggested that for most scales, the general cut-off point for an acceptable alpha should be .80 to reduce the effect of random measurement error. Thus, the alpha coefficients for all 58 items, as well as for each factor separately, were well within the acceptable alpha range. Even though acceptable alpha coefficients were achieved for the I 58 statements, an exploratory series of reliability analyses were 88 computed in an attempt to see if the item pool could be further reduced without reducing the alpha level below .80. Consequently, additional reliability analyses were computed for the 34 statements which had significantly discriminated between the CPE criterion groups 1 and 2 versus 4 and 5, and for the 56 statements which had significantly correlated with the CPE ratings. A summary of the alpha coefficients for each set of statements can be reviewed in Table 16. Table 16. Cronbach's Alpha Coefficients* for STQ-2 Items Selected on the Basis of Discriminant and Spearman Rank Order Analyses Factorsa Method of Item Selection I & II I II Alpha No.c Alpha No. Alpha No. Discriminantb .92 34 .94 28 .73 6 Correlational .94 56 .95 43 .85 13 8Factor I represents the Inappropriate items, Factor II represents the Appropriate items. bBased on CPE criterion groups of 1 and 2 versus 4 and 5. cNumber of items analyzed. 8p < .0001. An examination of the internal consistency of the pool of 34 statements, derived from the discriminant analysis, indicated that the Inappropriate Factor, as well as the combinationcfifFactors I and II, yielded sufficiently'highalpha coefficients of .94 and .92, 89 respectively. The Appropriate Factor, however, which contained only six items, yielded an unacceptable alpha coefficient of .73. The alpha coefficients based on the pool of 56 statements were very similar to those produced by the original analysis of 58 state- ments. The alpha coefficient of .95 for Factor II was only .01 lower than that derived from the original analysis. Furthermore, Factor I, as well as the combination of Factors I and II, produced alpha coefficients identical to those yielded by the analyses based on the pool of 58 items (see Table 16). The results of the reliability analyses calculated for the pool of 58 items were compared to the set of parallel analyses computed for the pool of 34 statements. An examination of the Inappropriate Factor coefficients for the two analyses (see Tables 15 and 16), revealed that a reduction of 17 statements altered the alpha coeffi- cient by only .02. Hence, it was assumed that a slight reduction of items within Factor I could most likely be tolerated without greatly decreasing Cronbach's alpha. Thus, six items which had not signifi- cantly discriminated between either of the CPE criterion groupings (CPE scores of 1 and 2 versus 5; and, scores of 1 and 2 versus 4 and 5) were subsequently eliminated, even though they significantly correlated (p < .05) with the CPE scores. Therefore, items were retained within the Ineffective Factor only if they significantly discriminated (p < .05) between one or both of the CPE groupings. This decision was based on the assumption that the correlational analysis appeared to be a more liberal test of group membership than 90 the discriminant analysis. Consequently, it seemed more apprOpriate to eliminate items from the group of statements which had a signifi- cant relationship with CPE, rather than from those which had signifi- cantly discriminated between the criterion groups. A similar comparison of Appropriate Factor alpha coefficients indicated that a reduction of items within this factor would be unwise, since the highest alpha based on the original analysis with 14 state- ments, was only .85. An addition of statements within this factor was not practical since the 14 items already represented all those statements which had significantly discriminated between, and/or correlated with the CPE criterion groups. Consequently, 38 items from the Inappropriate Factor and 14 items from the Appropriate Factor were retained, thereby reducing the total item pool from 58 to 52 statements. Cronbach's alpha was recalculated for the new pool of 52 state- ments. The results, as can be seen in Table 17, revealed that Cron- bach's alpha for all 52 statements combined was .93. The Inappropriate Factor yielded an alpha coefficient of .95, whereas the Appropriate Factor yielded an .85 alpha coefficient. The examination of the alpha coefficients for each of the 52 statements indicated that the omission of one of the items within Factor I, 34 ("I need a break") would increase the alpha coefficient more so than would any other item (although the rounded off value for alpha, with this item omitted would still equal .95). Furthermore, the statement, "I need a break" was a potentially confusing item. On one hand, it could be interpreted to mean that the person needed a 91 Table 17. Cronbach's Alpha Coefficients* for Remaining 52 STQ-2 Statements Factorsa I 8 II I II Alpha No .b Alpha No . Alpha No. .93 52 .95 38 .85 14 8Factor I represents the Inappropriate items; Factor II represents the Appropriate items. bNumber of items analyzed. {p < .0001. service break (to win a game that the opponent was serving), or it could be interpreted to mean that the player was wishing for a lucky turn of events. Hence, this statement was omitted from the item pool, thereby leaving a total of 51 statements to comprise the STQ-3. A final check of the alpha coefficients for the final 51 items, indi- cated that the rounded off coefficients were identical to those yielded for the analysis computed on the 52 items. Construction of STQ-3 On the basis of the foregoing analyses, the investigator assumed that the validity and reliability of the STQ-2 was adequate. That is, the process of retaining STQ-2 items which significantly discriminated between the CPE criterion groups, satisfied the requirements for con- struct validity (Carmines 8 Zeller, 1979). In addition, the selection of items which significantly correlated with the CPE ratings signified 92 criterion validity. Furthermore, alpha coefficients for all 51 items, as well as for each factor assessed separately, were all within the acceptable range of .80 and above, thereby establishing adequate instrument reliability. Since satisfactory levels of STQ-2 validity and reliability were established, construction of the final STQ-3 instrument was initiated. Table 18 gives a breakdown of Inappropriate and Appropriate statements within each of the four categories, for the 51 STQ-2 items selected to represent the STQ-3. As can be seen from Table 18, there were only three to five Appropriate statements per game situation category. Thus, in order to avoid clusters of these items within a category, the investigator randomly interspersed the Appropriate statements with the InapprOpriate statements by means of the table of random numbers. The sequential order of the Inappropriate Factor statements employed for the STQ-2 was retained for the STQ-3. The general format of the STQ-3 was identical to that of the STQ-2. Therefore, the direction page and personal data form, includ- ing skill andCHHEscales, were identical for both questionnaires. A copy of the STQ-3 can be found in Appendix I. Selection of Sample A variety of tennis players with different skill levels, degrees of competitive psychological effectiveness, ages, and years of playing experience, from lower Michigan, participated in the final phase of this investigation. Eligibility requirements for the STQ-3 partici- pation remained identical to those requirements mandated for STQ-2 93 Table 18.--Distribution of STQ-3 InapprOpriate and Appropriate Items Game Situation a Category Factor Item Total Leading I 8 II __5_ 13 Behind I 15 II _3 18 Pleased I II _ 9 Displeased I II I__ 11 Grand Total 15 a Factor I represents the Inappropriate items; Factor II represents the Appropriate items. participation, with one additional provision: that individuals had not participated in either of the first two phases of this investiga- tion. Again, as with the selection of players for the STQ-2, as many subjects as possible were asked to participate, so that the number of players minimally equaled the total number of statements (51) plus 94 two (Klecka, 1980). Specifically, an absolute minimum of 53 subjects had to be solicited to participate in this final phase. Administration of the STQ-3 Approximately half of the coaches/instructors whose involvement was solicited for Phase III, had also participated in the second phase of this investigation. The administrative procedures utilized for the STQ-2 were identical to those used for the STQ-3, with the exception of one important addendum. Each coach/instructor was asked to remind their players to respond to every item. Special emphasis on this point was mandated since a number of STQ-2 questionnaires had been discarded because of multiple missing responses. Coaches/ instructors were also asked to skim their players' questionnaires when they were returned, to ensure that their players had responded to every item, including all items on the player information page. All coaches/instructors who agreed to participate in this third phase insisted upon administering the STQ-3 themselves, at their own convenience. It was necessary, therefore, to once again use elaborate follow-up measures including multiple phone calls, written reminders, and personal visits, to encourage coaches/instructors to follow through with the administration and subsequent collection of the STQ-3. Extensive follow-up procedures were utilized in an attempt to achieve a recommended response rate of at least 50% (Babbie, 1979). However, of the 294 STQ-3 instruments which were distributed, only 98 (33.33%) were returned to the investigator. Only one STQ-3 had 95 to be discarded because of multiple missing responses. Therefore, 97 questionnaires were usable for the final phase of data analysis. Since there were no statistically based formulas for determining an appropriate response rate (Babbie, 1979), and since an adequate sample size had been attained, the decision was made to base the preceding statistical analyses on the 97 acquired questionnaires. Collection of Player NTRP and CPE Ratings As in Phase II, coaches/instructors were asked to classify each player according to the NTRP and CPE scales (see Appendices B and G). The results of NTRP and CPE scales, and the responses to the STQ-3 are presented in the following chapter. CHAPTER IV FINDINGS AND DISCUSSION Chapter IV will encompass the analysis and subsequent discussion of the findings from the administration of the STQ-3. Additionally, the results of the STQ-2 will be reviewed and compared to the STQ-3. Demographics of the STQ-3 Sample The demographic information pertaining to the subjects who com- pleted the STQ-3 is summarized in Tables 19 and 20. Additionally, Table 21 displays a comparative overview of the demographic data from all three STQ phases. A total of 97 subjects, 50 females and 47 males, participated in this final phase. Ages ranged from 16 to 62, with a mean age of 32.28. The five types of players represented in the Phase III sample included juniors, traveling team members, tennis professionals, college class students, and college team players. Two-thirds of these players were classified as traveling team members. The skill level of the Phase III subjects ranged from 3.0 to 7.0 (as measured by the NTRP) with 64% of the subjects rated as a 3.0 or 3.5. CPE and NTRP Ratings of the STQ-3 Sample All five levels of CPE and all nine skill levels were represented in the STQ-3 sample. The average CPE rating for these players was 96 97 Table 19. STQ-3 Frequencies and Percentages for Gender and Player Type Gender f % Player Type f % Female 50 52 College Class 11 11 College Team 14 14 Male 47 48 Junior 5 5 Professional 3 3 Traveling Team 64 66 Table 20. Descriptive Statistics on STQ-3 Personal Player Data Mean Range Median Mode SD Agea 32.28 16-65 33.00 18.0 11.80 Hoursb 4.67 1-20 3.57 2.0 3.30 Yearsc 7.62 1-30 5.58 5 . o 6 .39 CPE 3.31 1-5 3.29 3.0 .84 NTRP 3085 300—700 3053 300 093 aAge B player's age. bHours - number of hours of tennis played per week. cYears = number of years having played tennis. Table 21. Comparison of Descriptive Statistics on Personal Player Data for the STQ-l (N - 54), STQ-2 (N = 151), and STQ-3 (N = 97) Mean Range Median Mode S.D. Agea STQ-1 26.13 16-59 21.17 19.0 10.90 STQ-2 31.74 16-55 33.86 16.0 11.70 STQ-3 32.28 16-65 33.00 18.0 11.80 Hoursb STQ-1 5.98 1-50 3.42 2.0 8.60 STQ-3 4.67 1—20 3.57 2.0 3.30 Yearsc STQ-1 7.68 1-50 3.25 1.0 10.30 STQ-2 7.48 1-30 5.30 5.0 5.94 STQ-3 7.62 1-30 5.58 5.0 6.39 989 STQ-2 3.35 1-5 3.44 4.0 .91 STQ-3 3.31 1-5 3.29 3.0 .84 NTRP STQ-2 4.29 3.0-7.0 4.01 3.5 1.08 STQ-3 3.85 3.0-7.0 3.53 3.0 .93 NOTE: Coaches were not requested to assign CPE or NTRP ratings for the STQ-1 participants. aAge = player's age. b cYears = number of years having played tennis. Hours = number of hours of tennis played per week. 99 3.35, which was slightly higher than the mean score of 3.31 for the STQ-2 athletes. The average coach assigned NTRP score was 3.85, as compared to a mean NTRP score of 4.29 for the STQ-2 players. Tables 22 and 23 given the STQ-2 and STQ-3 CPE and NTRP score distributions, respectively. Table 22. STQ-2 and STQ-3 CPE Score Distribution CPE STQ-2 STQ-3 Score f % f % l 6 4 2 2 2 17 11 11 11 3 56 37 45 46 4 62 41 33 34 5 10 7 6 6 N = 151 N = 97 M = 3.35 M = 3.31 100 Table 23. STQ-2 and STQ-3 NTRP Score Distributions NTRP STQ-2 STQ-3 Score f % f % 3.0 26 17 31 32 3.5 37 25 31 32 4.0 24 16 7 7 4.5 15 10 8 8 5.0 13 9 10 10 5.5 18 12 7 7 6.0 9 6 1 1 6.5 8 5 1 1 7.0 ‘__l 1 ___l_ 1 N 151 N = 97 M 4.29 M = 3.85 SD 1.08 SD = .93 101 Paychometric Properties of the STQ-3 Construct Validity The development of a new instrument must always include a determination of the extent of its validity. The validity of the STQ-3 was assessed in a number of different ways. Discriminant function analysis. First, a discriminant function analysis was conducted to determine whether or not the player responses to the 51 STQ-3 items significantly discriminated between the criterion groups of psychologically effective (CPE assigned ratings of 4 or 5; .n_= 39) and psychologically ineffective (CPE assigned ratings of 1 or 2;.n = 13) players. Only 50 of the 51 statements were processed for the discriminant function analysis. Item D-9 ("I feel foolish. Why don't I concentrate") was deleted from this analysis because it failed the tolerance test. This statement yielded a tolerance of zero which indicated that it was a linear combination of one or more of the STQ-3 items (Klecka, 1980). Since the two criterion groups of psycho- .logically effective and ineffective players comprised a total of 52 subjects, and since the number of statements analyzed had been reduced ‘to 50, the requirements for a necessary and sufficient 2 (two greater tflnan the number of discriminating variables) were met. The null hypothesis which assumed that the mean vectors between ‘the two CPE groups were equal was examined by means of the Wilks' lambda criterion. Since Wilks' lambda represents the percent of ‘unexplained variance, low values were preferred. Values for Wilks' lambda ranged from a low of .43090 (p < .00001) for statement L5 102 (I'm ahead now, but it probably won't last") to a high of .99990 UR < .9436) for statement D2 ("You're catching it late"). A com- plete listing of the Wilks' lambda values for the 50 STQ-3 items included in the discriminant analysis is given in the table located in Appendix J. Thirty-four of the 50 STQ-3 items analyzed, significantly dis- criminated (p < .05) between the two criterion groups of psychologi- cally effective and ineffective players. Thirty-two of these 34 statements were those which had previously been described as Inap- pr0priate statements (Factor I items). Only two Appropriate statements (Factor II items) significantly discriminated (p_< .05) 'between the two criterion groups. A pictorial display of the unique 'response patterns of the psychologically ineffective and effective ‘players to the 32 Inappropriate Factor discriminating items is given in Figure 2. As expected, the psychologically ineffective players :responded with a higher frequency to the Inappropriate statements ‘than.did the psychologically effective players. Similarly, the liatter group responded with a higher frequency to the two discrimin- ating Appropriate items than did the psychologically ineffective Players did . A total of 16 STQ-3 statements failed to discriminate signifi- Cantly (p > .05) between the designated CPE criterion groups. Nelve (If these 16 statements were from the Appropriate Item Factor. Further- DMJre, 7 of the 16 statements had also yielded nonsignificant results 111 the parallel STQ-2 discriminant analysis. These seven items were .meoamfim m>wuommem new o>HuoowmwcH zHHmofiwoaooohmm an macaw mlcem oumprOLoamcH «wcfiuocfiefiuom«c mm Cu cuouomm uncommom .N ouomfia wmoome n x, 1 x a m m >1. , a TI! ~’ I \/ sloéw \\‘ \ ’ x z \\\\)r \ / s m I a \ /\ a «x I \ r I a I \ /\ K x 1% J ‘1 m / ‘0 S “re J xex 1.1L 4.... Is «SSH V a. N mo.vm. 19m muozmaa m>auoommm an mouoom unoeoumum coo: muoamao o>auoowwoow an mouoom acosmumum :moz 1111 104 the only STQ-2 items which had not significantly discriminated between the CPE criterion groups, but had been retained for the STQ-3 to increase the reliability of the Appropriate Factor statements to an acceptable level. In summary, then, a total of nine statements which had discriminated significantly between the STQ-2 criterion groups, did not significantly discriminate between the STQ-3 CPE criterion groups. Since 34 of the 50 items significantly discrimin- ated between both the STQ-2 and STQ-3 psychologically effective and ineffective player groups, the STQ-3 was judged to have construct validity. Discriminant classification analysis. The construct validity of the STQ-3 was additionally assessed by means of a discriminant classi- fication analysis. This procedure predicted group membership for each participant into either the psychologically effective or ineffective category, based on that player's scale score for the entire STQ-3. The results revealed that when the predicted group membership was compared to the actual membership, the classification process was 100% accurate. These results reflected the strength of the Appropriate and Inappropriate statements as discriminating variables, and thereby further supported the construct validity of the STQ-3. Factor analysis. A factor analysis was conducted to enable the investigator to discern if the underlying constructs of the STQ-3 responses were the same or similar to those formed by the STQ-2 responses. A principle-components factor analysis with iterations 105 was once again utilized. Similar to the analysis of the STQ-2 responses, 14 factors emerged with eigenvalues greater than 1.00. Each of these 14 factors was scrutinized for its potential contribu- tion to the number of underlying constructs. Once again, the con- structs for Factors III through XIV were nonsensical and were thereby omitted from further analyses. The factor loadings computed for each statement within the two remaining factors were then examined. The assignment of the STQ-3 items to Factor I or II was identical to that of the previous factor analysis based on the STQ-2 data. That is, Factor I was again clearly characterized by Inappropriate statements, including self-critical, low self-efficacy, and off-task statements. Factor II was characterized by Appropriate statements, including on-task statements/self-directions, and words of encouragement and praise. The combination of these two factors accounted for 34.9% of the total variance. Factors I and II each accounted for 25.3% and 9.6% of the variance, respectively. These values were similar to those found for the STQ-2 as Table 24 indicates. Table 24. A Comparison of STQ-2 and STQ-3 Percent Variance and Eigenvalues for Factors I and II Percent Variance .Eigenvalue State- ment Inappropriate ApprOpriate Cum. Inappropriate Appropriate Factor I Factor II % Factor I Factor II STQ-3 25.3 9.6 34.9 12.9 4.9 106 The similarities between the STQ-2 and STQ-3 factor analysis results, including the identical assignment of items into Factor I or Factor II, further supported the construct validity of the STQ-3. Reliability Cronbach's alpha was once again utilized to assess the degree of reliability for the STQ-3. The results of the reliability analysis on the STQ-3 items, in addition to a comparison of the STQ-2 and STQ-3 alpha coefficients, are given in Table 25. Table 25. Cronbach's Alpha Coefficients* for Final 51 Items Based on STQ-2 and STQ-3 Data Factorsa Instrument I 8 II I II Alpha No.b Alpha No. Alpha No. STQ-3 .92 51 .94 37 .81 14 3Factor I represents the Inappropriate items. Factor II represents the Appropriate items. bNumber of items analyzed. 5p < .0001. STQ-3 alpha coefficients of .92 for all 51 items, and .94 and .81 (p < .0001) for the Inappropriate and Appropriate Factor items, respectively, were slightly lower than those based on STQ-2 data. 107 The alpha coefficients attained for both factors together and separately fell above the acceptable alpha minimum of .80 (Carmines & Zeller, 1979). The STQ-3 was, therefore, judged to be reliable in its entirety, as well as for each factor separately. Furthermore, the similarity of the STQ-2 and STQ-3 alpha coefficients supported the stability of the STQ-3. Analysis of Results Thus far the data have been used in analyses which determined the degree of validity and reliability of the STQ-3. These analyses revealed two important aspects regarding the nature of competition- related mental monologues. First, results of the factor analysis indicated that such self-talk could be classified into Inappropriate items and Appropriate items. Second, 94.28% of the statements which discriminated significantly between the CPE criterion groups were Inappropriate items. These findings were subsequently supplemented with an analysis of mean scores and standard deviations for STQ-3 responses by all players. Similar to the player statement trends revealed by the discriminant function analysis, the mean scores reinforced the same trends for the CPE groups as Table 26 illustrates. Thus, an examina— tion of the response pattern for the Inappropriate items revealed an inverse relationship, so that as psychological effectiveness increased the frequency with which players identified with the InapprOpriate statements decreased. Similarly, an examination of the Appropriate items seemed to suggest a direct relationship between the players' 108 Table 26. Means and Standard Deviations for Player Responses to Inappropriate and Appropriate Factor Items a Inappropriate Items Appropriate Items CPE Groups N M S.D. M S.D. 4 or 5 39 2.11 .51 3.39 .74 3 45 2.56 .53 3.28 .44 1 or 2 13 3.22 .38 3.08 .53 aIndividuals assigned a CPE rating of 4 or 5 represented the psy- chologically effective players, whereas those assigned a CPE rating of 1 or 2 represented the psychologically ineffective players. CPE level and response rate to the Appropriate statements. That is, as players' CPE level increased, the frequency with which they iden- tified with the Appropriate items also seemed to increase. The data were subsequently subjected to a multivariate analysis so that the significance of these trends and apparent self-talk differences between the CPE criterion groups could be tested. A multivariate design was selected, instead of independent analyses on each variable, to decrease the chances of Type I errors, and to maximize the amount of information gained through the interrelation- ships of the Inappropriate and Appropriate Factor items (Schultz, Smoll, & Gessaroli, 1983). A one-way multivariate analysis of covariance (MANCOVA) design was utilized, which employed two dependent variables,one independent variable, and two covariates. The scale scores for both the 109 Inappropriate and Appropriate Factor items served as the dependent variables. The independent variable was represented by two extreme CPE groupings of players, the psychologically effective (players with CPE ratings of 4 or 5) and ineffective players (CPE ratings of 1 or 2). The two covariates selected for this analysis were gender and skill (NTRP). These two concomitant variables were incorporated into the statistical design to help control for possible experimental varia- bility (Winer, 1971). Gender and skill were conceptually chosen over all other possible confounding variables assessed by the STQ-3, including age, years of playing experience, hours of tennis played per week, and player type, because of a potential gender/skill bias which could have affected the results. That is, females who completed the STQ-3 were lower skilled on the average Qfl = 3.56) than the males Q5 = 4.16). In fact, 84% of the females, as compared to 43% of the males, had received a NTRP score of 3.0 or 3.5. Given that the sex by skill distribution was skewed, these two covariates were chosen then to minimize the possible confounding effects they might individu- ally or mutually have had on the response patterns from the various CPE comparison groups. Furthermore, the inclusion of these concomi- tant variables in the multivariate design may have increased the power of the test by increasing the sensitivity of the statistical procedure to the differences among the group assessed (Huck et al., 1974). The null hypothesis, associated with the onedway MANCOVA design selected for this investigation, assumed that the Inappropriate and 110 Appropriate Factor item mean vectors would not differ between the two CPE criterion groups. The Wilks' lambda criterion was utilized to assess the probability level of attaining equal mean vectors, as mandated by the null hypothesis (Cooley & Lohnes, 1971). Wilks' lambda yielded a value of .73840 (p < .000001), and thus the null hypothesis was rejected. The use of Wilks' lambda, however, is based on the assumption that the matrices of a given population are equally dispersed (Cooley & Lohnes, 1971; Hays, 1973; Winer, 1971). The Box's M test utilized to test this assumption, yielded highly significant results (F_= 2.74 (20,4945), 2 < .00006), thereby violating the assumption of homogeneous dispersion matrices. The seriousness of this violation is questionable however, since many contemporary statisticians have agreed that the importance of homogeneity has been overrated in the past (Cooley & Lohnes, 1971; Hays, 1973; Kerlinger, 1973; Winer, 1971). Several statisticians (Cooley & Lohnes, 1971; Hays, 1973; Kerlinger, 1973) have stated that the MANOVA/MANCOVA statistical pro— cedures are powerful enough to withstand departures from the assump- tion of homogeneous dispersion matrices. Furthermore, Hays (1973) suggested that due to the robustness of the MANOVA/MANCOVA pro- cedures, investigators could employ these analyses without specifi- cally testing for homogeneity. On the basis of these recommendations, the investigator con- cluded that the results of the MANCOVA procedure could be explored. If, in the current study, the results had been affected by the hetero- geneity of variance, at worst, the value of Wilks' lambda could have 111 been inflated. Since the Wilks' lambda criterion was overwhelmingly significant (p_< .000001), however, there appeared to be very little chance of making a Type I error. Thus, the MANCOVA results were examined with the understanding that the significance level of the Wilks' lambda criterion might possibly be somewhat lower than the value yielded, although not so low as to fail to reject the null hypothesis. The rejection of the null hypothesis thereby indicated that the groups of psychologically effective and ineffective players were sig- nificantly different (p < .0000001) from each other in terms of their mean scores on the STQ-3. In other words, the STQ-3 accomplished what it was designed to do, that is, statistically differentiate between the psychologically effective and ineffective players on the basis of their STQ-3 self-talk responses. A follow-up test to the Wilks' lambda criterion was needed to assess whether the Appropriate and/or Inappropriate Factor items were contributing to the significant results. A discriminant function analysis was thereby employed. The results of this analysis are given in Table 27., The standardized, rather than the raw discriminant function coef- ficients, were examined so that the means and standard deviations of the ApprOpriate and Inappropriate Factor items could be adjusted to account for the difference in number of items in Factor I (§_= 37) and Factor II (§_= 14). As can be seen from Table 27, the standardized coefficient of 1.01385 for the Inappropriate Factor was substantially higher than that of -.37106 for the Appropriate Factor. Furthermore, the univariate F value of 28.34270, representing the Inappropriate 112 Table 27. Discriminant Function Results for Inappropriate and Appropriate Factor Items Source of MS Between Univariate b Variation Variable Groups Fa SDFC CPE Inappropriate 6.57169 28.34270** 1.01385 Criterion Factor Items groupsc Appropriate .24103 .70813* -.37016 Factor Items adf = 1,92. bSDFC = Standardized discriminant function coefficients. cGroups were controlled for the effects of sex and skill. {p < .402. *fp < .0000007. statements, was highly significant (p < .0000007), whereas the uni- variate F value of .70813 for the Appropriate statements, was not significant (p < .402). Hence, on the basis of the results of the discriminant function analysis and the univariate F-tests, the Inappropriate Factor statements were judged to be the major con— tributing component of the significant Wilks' lambda criterion. The ability of the Inappropriate statements to significantly differentiate among the various levels of CPE is pictorally displayed in Figure 3, where the 95% confidence intervals of the item responses for three CPE levels have been plotted. Even though mean score trends existed for CPE group responses to both the InapprOpriate and Appro— priate items, the confidence intervals clearly exhibit the distinct CPE group response patterns for the InapprOpriate items, and the overlapping response patterns for the Appropriate statements. 113 5 F' Inappropriate Item Appropriate Item Responses Responses E 4- o S I “5g 3_l I I 38 I c o. 33 2.. I own m Li a. 1 l 1 1 l I J 1 & 2 3 4 & 5 1 & 2 3 4 & 5 L Players Grouped by CPE Ratings Figure 3. STQ-3 95% Confidence Intervals for CPE Group Responses. Summary of Results It is apparent from the findings discussed thus far that the STQ-3 was capable of measuring competition related self-talk, and that certain patterns of self-talk distinguished the psychologically effective players from the ineffective players. Furthermore, it was discovered that it was mainly the responses to the Inappropriate Factor items which significantly differentiated the psychologically effective from the ineffective players. This phenomenon was prevalent throughout the construction and subsequent administration of the STQ-3. For instance, the factor structure for both the STQ-2 and STQ-3 revealed that Factor I was consistently represented by the Inappropriate statements, and thereby accounted for more variance 114 than Factor II. Likewise, the results of the Phase II and Phase III discriminant analyses revealed that 84.44% and 94.28%, respectively, of the discriminating statements were Inappropriate Factor items. Furthermore, reliability analyses computed for both the STQ-2 and STQ-3 indicated that the Inappropriate Factor items had yielded the highest reliability coefficients, even higher than the coefficients computed for the instrument as a whole. Finally, the discriminant function analysis and univariate F-test computed as a follow up to the significant Wilks' lambda (p < .0000001) derived from the STQ-3 MANCOVA analysis, revealed that the Inappropriate items were the major contributing source to the highly significant results obtained. Discussion and Implications The samples of tennis players chosen for the various phases of this investigation appeared to emit two basic types of self- statements during a competitive event. Statements such as on-task expressions/self-directions, and words of encouragement and praise were labeled as the Appropriate self-statements. Conversely, the self-critical, low self-efficacy, and off-task phrases were labeled as the Inappropriate items. The investigations reviewed in Chapter II, which focused on the assessment of an individual's self-talk content, consistently revealed that individuals who displayed a particular maladaptive behavior, also emitted a relatively high frequency of corresponding Inappropriate/Maladaptive self-statements. Some of the maladaptive behaviors assessed included: high social anxiousness (Cacioppo et al., 1979); high test anxiousness 115 (Hollandsworth et al., 1979); depression (Hollon & Kendall, 1980); poor coping ability for a medical procedure (Kendall et al., 1979); and low assertiveness (Schwartz & Gottman, 1976). In the current investigation the effective players emitted mostly Appropriate self-statements while competing, whereas the ineffective players engaged in a moderately high frequency of both Inappropriate and_Appr0priate self-statements. Consequently, when the ineffective players were competing in a tennis match, they may have been emitting an array of conflicting cognitions. Thus, these hypothesized clashing cognitions may have been a contributing factor to the ineffective players' maladaptive performance. When Kendall and Hollon (1981) reviewed the available self—talk assessment literature, including their own investigations, they too found that individuals who displayed certain maladaptive behaviors also voiced a mixture of both Appropriate and Inappropriate self- statements. Hence, they theorized that the absence or omission of negative/maladaptive cognitions may be more influential in promoting an adaptive behavior than the emission of positive/adaptive self- talk. They referred to this hypothesis as the "power of non- negative thinking" (p. 110). This theory is different from the concepts governing the power of positive thinking. The focus of the "power of non-negative thinking" is on the omission of the Inappro- priate statements, rather than on the emission of the ApprOpriate phrases. Contemporary self-help books and articles are replete with examples extolling the use of positive/adaptive self—talk to enhance 116 one's behavior/performance (Gologor, 1979; Spinelli & Barrios, 1980; Kirschenbaum & Bale, 1980; Kauss, 1980; Harris & Harris, 1984; Rushall, 1979; Orlick, 1980). Many of these authors have suggested that behavioral adaptations and/or performance improvements will occur if individuals repeat positive/adaptive statements to themselves before and/or during the designated behavior. Perhaps the notion that induced positive self-talk would aid in behavioral adaptations was based in part on the assumption that if negative self—talk is associated with maladaptive behavior, and if positive self-talk is associated with adaptive behavior, then individuals should be encouraged to voice positive self-statements. However, as Meichen— baum (1977) pointed out, the mere parroting of positive self-statements might be somewhat helpful during low anxious situations, but the effectiveness of the given platitudes often diminish and/or completely disappear during high stress periods. This point was reinforced by Girodo and Roehl (1978) who found that the emission of positive/coping self statements did not significantly alter a particular maladaptive behavior under high stress conditions. In fact, the investigators suggested that induced self-talk may even act as a distractor and consequently hinder effective coping behaviors. It is possible, then, that the theorem of "non-negative thinking" may restructure the way we have conceptualized the effects of positive self-talk on adaptive behavior, as well as on behavior change. Future efforts to modify or eliminate maladaptive behavior might include the identification of Inappropriate self-talk patterns, as well as methods 117 to minimize or eliminate those potentially disruptive cognitions, rather than encouraging the client/athlete to emit Appropriate self- statements. Before such recommendations and global implications are made, however, much systematic research is still needed to establish causality as well as to substantiate or refute the advantages of eliminating Inappropriate self-talk patterns. Although the results of this investigation are thought provoking and encouraging in that a particular cognitive pattern emerged to distinguish the psychologically effective from the ineffective players, it is important to recognize that there is no evidence at this point to declare that the inconsistent performance displayed by the ineffective players was actually caused by Inappropriate self-talk. While it is possible that Inappropriate monologues could have had an adverse effect on performance, it is just as likely that a player's performance could also have been adversely affected by any one or combination of player "pathologies" including fear of failure, fear of success, uncontrollable levels of game related anxiety, poor concentration skills, differential levels of motivation, improper attributional styles, low self-efficacy, etc. It is also possible that Inappropriate self-talk patterns and the aforementioned player "pathologies" could have been interrelated. That is, perhaps in some cases the Inappropriate self-statements were only symptoms of deeper and more complex problems, such as the fear of failure or success. In such cases, there would be a need for a more comprehensive therapeutic approach to alter a player's performance, as opposed to 118 simply restructuring and/or eliminating the individual's Inappro- priate self-talk. Regardless of whether the Inappropriate self-statements are merely a symptom of deeper problems, or if they are directly contribu- ting to decrements in performance, a numbercfifquestions need to be addressed before sensible and responsible cognitive-behavior modifi— cation programs are implemented. 1. Are there individuals who perform erratically during competitive events, but do not simultaneously engage in Inappropriate self-talk? Are there individuals who perform consistently up to their potential during competitive events, but yet engage in a relatively high percentage of concurrent Inappropriate self-talk? If an individual engages in an Inappropriate mono- logue while playing inconsistently, will the elimina- tion of the Inappropriate statements lead to a more consistent performance? If an individual plays consistently up to his/her potential during a competitive event, and is found to engage in situation Appropriate self-talk, will a decrement in performance occur if, through experimental manipulation, the player switches to a monologue of Inappropriate self—talk? Should situation-Appropriate self—talk statements be substituted at the times/places players formerly voiced 119 Inappropriate phrases? Or, would it be more effective to encourage athletes to quiet their minds at the times/places when they formerly emitted the Inappro- priate phrases? As was stated in Chapter I, the investigator hoped that the STQ-3 could be used to assist in a few fundamental assessment func- tions, as outlined by Nideffer (1976). These functions included: 1. The identification of possible self-talk problem areas. 2. The enhancement of a coach's understanding of player behaviors. 3. The improvement of communication channels between the coach and player. It is apparent, however, that a number of important questions must be addressed, and perhaps subsequent instrument modification imple- mented before the existing STQ-3 can prudently be integrated into an effective cognitive behavioral modification program, and/or used to assist in the aforementioned assessment functions. In summary, then, while it appears that the development of the STQ-3 was a major step forward in the assessment and eventual com- prehension of the relationship between a player's self-talk and concurrent performance, it is not recommended that the STQ-3 be employed by coaches and teachers as an assessment tool, at this point in time. Rather, the employment of this instrument could be used as a stepping stone for future empirical investigations to further explore the relationships between self-talk and performance patterns. CHAPTER V SUMMARY, CRITIQUE, AND FUTURE DIRECTIONS Chapter V includes a review of the three major developmental phases of the STQ-3. Additionally, the chapter contains a summary of the findings from the data analyses on all three STQ instruments. The implications of this investigation, and suggestions for future research are also discussed. Summary The intent of this investigation was to design a self-talk ques- tionnaire (STQ-3) which could validly and reliably assess the type and frequency of tennis players' competition related self-talk as it pertained to competitive psychological effectiveness. Three major investigative phases were implemented in the development of the final instrument, the STQ-3. Phase I was designed to amass a large pool of competition related self-statements generated by male and female intermediate to advanced tennis players. The investigator achieved this objective by administrating the STQ-1 to 54 players from Florida, Michigan, and Texas. Directions given for the STQ-l asked players to mentally image the last time they competed under each of four general game situations, including when they were ahead in a set, behind in a set, 120 121 pleased with their play, and displeased with their play, and then, to recall and record their game related self—talk during each of the given situations. This process aided in the generation of more than 2,000 competition related self-statements. In Phase II, the investigator selected a representative sample of statements from the item pool generated in Phase I, and used those statements to construct the STQ—2. The reduction of the Phase I item pool involved several steps. First, the entire pool was screened for redundant and/or inappropriate responses. This process reduced the item pool to 726 statements. Two independent judges then classi- fied these remaining statements into five categories derived from the Klinger et a1. (1981) investigation. This process enabled the judges to recognize and subsequently delete similar phrases within each of the categories, thereby reducing the item pool to 663. A final reduction method involved the employment of five independent raters who classified each of the 663 statements as being characteristic of either a psychologically effective or ineffective player, and as being characteristic of either a skilled or unskilled player. As a result, a total of 118 statements, which generated 100% rater agree- ment across both categories were retained. These items consequently comprised the STQ-2. The STQ-2 was designed to reflect the frequency with which players emitted a selected number of self-statements when competing under each of the four game situation categories of Leading-, Behind in a set, and Pleased—,Displeased with play. A S-point rating scale 122 ranging from 1 = Not at All, to 5 = Almost Always, was incorporated to measure self-talk frequency. In addition to the assessment of self-talk frequency, coaches rated each of their players according to the athlete's level of competitive psychological effectiveness (CPE). A 5-point scale was employed which ranged from 1 a rarely plays to top of potential, to 5 = almost always plays to top of poten- tial. Coaches rated each participating player on two separate occa- sions so that the reliability of the CPE scale could be determined. The STQ-2 was distributed to 407 intermediate to world class tennis players, ages 16 and older, from lower Michigan. A total of 179 questionnaires were returned (a 42% return rate) of which 151 were useable for subsequent statistical analyses. The third and final phase of the STQ-3 development included sta- tistical analyses of the STQ—2 responses and consequential systematic elimination of STQ—2 statements to create a refined item pool for the STQ-3. Since one of the major intentions of this investigation was to determine if the competition related self-talk emitted by psy- chologically effective and ineffective players differed significantly (p < .05), a series of discriminant analyses were employed to deter- mine which of the 118 STQ-2 statements differentiated between various CPE criterion groupings. Results indicated that a total of 45 items significantly discriminated (p < .05) between the designated CPE criterion groups. In addition to the discriminant analyses, a correlational analy- sis was also computed to serve as a supplemental piece of information 123 to the discriminant analyses. A total of 56 items was found to significantly correlate (p < .05) between levels of CPE and the item response. Since the STQ-3 was to be designed to detect significant differences in the type of self-talk emitted by the CPE criterion groups, the 60 items which neither significantly discriminated (p < .05) between the designated CPE criterion groups, nor signifi- cantly correlated (p < .05) among the CPE levels and item responses, were thereby eliminated. This reduced the item pool from 118 to 58 statements. The responses associated with the remaining 58 statements were further scrutinized by means of a principal components factor analy- sis. This analysis was employed to determine whether or not the statements could be grouped together by two or more specific under- lying constructs. The results revealed that the STQ-2 consisted of two well defined constructs, with Factor I accounting for 26.7% of the variance, and Factor 11 accounting for 9.5% of the variance. Factor I, labeled the Inappropriate Factor, was characterized by low self-efficiacy statements, including self-critical, and off-task statements. Conversely, Factor II, the Appropriate Factor, was char- acterized by on-task statements, including constructive self- directions, and words of encouragement and praise. The investigator also examined the extent of the instrument's reliability in addition to its validity. The data for the remaining 58 statements were, therefore, subjected to a series of reliability analyses, based on Cronbach's Alpha. Initially, Cronbach's alpha 124 was computed for all 58 items as a unit, and for each factor separately. Results yielded alpha coefficients of .94 for all 58 items, and .96, and .85 for Factors I and II, respectively. Since the alpha coeffi- cients attained were well within the acceptable range for instrument reliability, two additional reliability analyses were computed in an attempt to determine if the item pool could be further reduced without a substantial decrease in the alpha coefficients. Hence, Cronbach's Alpha was computed for the 34 items which had significantly discrimin- ated between the CPE criterion groups (based on ratings of 1 or 2, and 4 or 5), as well as for the 56 items which correlated significantly with the levels of CPE. On the basis of these results, it was discovered that a number of items within Factor I could be omitted without greatly reducing the alpha level. Therefore, six Factor I statements, which had not significantly discriminated (p_> .05) between either of the CPE criterion groups were eliminated, even though the responses to those six items had significantly correlated with the five CPE levels. Thus, the item pool was reduced to 52 statements. Finally, when all of the alpha coefficients for the remaining 52 statements were scrutin- ized, it was discovered that the omission of one final item, within Factor I, would increase the alpha coefficient more so than would the elimination of any other item. Furthermore, this particular item was potentially confusing, since two very different meanings could be construed from the same statement. Thus, that particular item was omitted, leaving a total of 51 statements to comprise the STQ-3. A final reliability analysis based on the 51 items yielded alpha 125 coefficients of .93 for the entire pool of staements, and .95 and .85 for Factors I and II, respectively. The format of the STQ-3 was identical to the STQ—2. The only difference between the two instruments was that there were fewer STQ-3 items presented within each of the four game situation cate- gories (Leading, Behind, Pleased, Displeased). The STQ-3 was subsequently distributed to 294 intermediate to world class players who had not previously participated in this inves- tigation. Additionally, coaches were asked to assign each of their players a skill rating, as well as a CPE rating. A total of 98 ques- tionnaires (33%) were returned of which 97 were usable for the final phase of data analyses. Once again, a discriminant function analysis was employed to determine if the responses to the 51 STQ-3 items significantly dis- criminated between the criterion groups of psychologically effective (E = 39) and psychologically ineffective (§’= 13) players. Results revealed that a total of 34 STQ-3 statements significantly discriminated between the two CPE criterion groups. Interestingly, 32 of the 34 items were those which had previously been described as Inappropriate (Factor I) statements. Thus, only two Appropriate (Factor II) items significantly discriminated between the crition groups. A discriminant classification analysis was also employed to determine the effectiveness of predicting CPE criterion group member- ship, based on the player's scale score on the STQ-3. Results revealed that the classification process was 100% accurate at predicting whether an individual was a member of the effective or ineffective criterion 126 group. These results added support to the construct validity of the STQ-3. The STQ-3 item responses were further scrutinized by means of a factor analysis to determine if the STQ-3 factor structure was the same or similar to that based on the STQ-2 item responses. Once again, a principal components factor analysis was used. Results indicated that the same two distinct underlying constructs prevailed. The com- bination of these two factors accounted for 34.9% of the variance. When viewed separately, Factor I accounted for 25.3%, and Factor 11 accounted for 9.6% of the variance. The assignment of STQ-3 statements into either Factor I or II was identical to the placement of the STQ-2 items. The similarities of the STQ-2 and STQ-3 factor analysis results added further support to the construct validity of the STQ-3. Next, Cronbach's alpha was employed to examine the extent of the instrument's reliability. Although the alpha coefficients of .92 for all items and .94 and .81 (p < .0001) for the Inappropriate and Appro- priate statements, respectively, were slightly lower than those coeffi- cients based on the STQ-2 item responses, the STQ-3 was judged to be reliable, since all coefficients fell within the acceptable alpha range for instrument reliability. Finally, the data were subjected to a one-way MANCOVA analysis to examine the significance of the self-talk trends and apparent differ- ences between the CPE criterion groups. A highly significant Wilks' lambda value of .73840 (p < .000001) resulted, thereby rejecting the null hypothesis of equal mean vectors. A subsequent examination of 127 the standardized discriminant function coefficients revealed that the value of 1.01385 for the Inappropriate Factor was substantially higher than that of -.37106 for the Appropriate Factor. Furthermore, the highly significant univariate F value of 28.34270 (2 < .0000007) for the Inappropriate items was appreciably higher than the nonsignificant univariate F value of .70813 (p < .402), thus indicating that the Inappropriate items were the major contributing component of the significant Wilks' lambda criterion. In summary, then, based on the battery of analyses computed on the STQ-3 item responses, the STQ-3 was judged to have adequate instrument validity and reliability. Furthermore, it was discovered that the responses to the Inappropriate statements consistently accounted for the vast majority of significant self-talk differences between the psychologically effective and ineffective criterion groups. Hence, the overall investigative effort was judged to be successful and worthy of future systematic study. Critique of STQ-3 and Future Directions Instrument Variation While the current investigative STQ-3 findings have supported its validity and reliability, the STQ-3 is not problem free. For instance, a factor analysis computed on the STQ-3 player responses revealed that only 34.9% of the total variance was accounted for by the combination of Appropriate and Inappropriate Factor items. Hence, approximately 65% of the instruments total variance was unexplained. Since the response patterns were consistent from Phase II to Phase III, 128 it does not seem likely that the large portion of unexplained variance could have been totally attributed to sampling methodology. Further- more, since the core of items which comprised the STQ-3 was systemati- cally derived from the original pool of self-Statements collected during Phase I, it seems improbable that 65% of the variance could have been accounted for by a misrepresentative sample of self-statements. In addition, the amount of variance accounted for by both the Appro— priate and Inappropriate Factors ranged from 30% based on the 118 STQ-2 items to 34.9% for the 51 STQ-3 items. Thus, it does not appear as though the reduction in the number of statements could have been solely responsible for the unexplained variance. The investigator sought a plausible explanation for the unex- plained instrument variance, by comparing the results of this inves- tigation to the findings based on Hollon and Kendall's (1980) Automatic Thoughts Questionnaire (ATQ-30). The development of the STQ-3 was patterned after the ATQ-30, thus the results of a factor analysis, computed on subject responses to the ATQ-30 items, was examined to determine if similar variance problems had been encountered. In contrast to the STQ-3 factor analysis results, the ATQ-30 accounted for 58.9% of the total variance, with Factor I accounting for 45.9% of the variance. The difference between the STQ-3 and ATQ-30 factor analysis results might have been due to differences in the subject population utilized, and the type of problematic behavior assessed. The ATQ-30 was designed to distinguish between psychometrically defined groups of depressed and nondepressed individuals on the basis 129 of their depression related self-statements. It is possible that the self-talk emitted by a clinically depressed sample of subjects could have been more homogeneous and might have exemplified a narrower focus than the self-talk emitted by a "normal" sample of tennis players with a conceivably broader cognitive focus. If this contention were true, a larger sampling error for the STQ-3 may have occurred, due to the relatively heterogeneous nature of the item pool, which, in turn, may have increased the degree of STQ-3 variability. Another possible source of response variability within the STQ-3 may have been due to the administrative flexibility permitted. Since coaches/instructors insisted upon administering the STQ-3 at their own convenience, the investigator had very little control over the uniformity of the administrative procedures. Thus, some players completed the STQ-3 at home, some answered it immediately following match play, while others responded to it prior to a practice period or a competitive match. In addition, a player's motivation to care- fully complete the STQ-3 may have in part been influenced by the coach/instructor's presentation of the investigation and subsequent administration of the STQ-3. Furthermore, certain players may have felt the need to censor their responses, since in most cases they were asked to return the completed questionnaire to their coach/ instructor, thereby creating the potential for added instrument vari— ability. Although the possible sources of instrument variance suggested, including environmental distractors, time restraints, differential 130 motivation, and censorship, are not inclusive, any one or combination of the ideas presented may have accounted for at least a portion of the unexplained variance. Concomitant research is needed to examine if additional variation could be explained by a systematic control of one or more of the potential problem areas previously implicated. CPE Design Difficulties In addition to the perplexity of the unexplained portion of STQ-3 variance, the format of the CPE scale demonstrated inherent design difficulties. Even though there were no measurement problems associated with the scale per se, and it clearly measured what it was designed to measure, a critical examination of the players' CPE fre- quency distribution revealed major problems with the basic design of the scale. For instance, there were very few players who were rated as those who rarely played to the top of their potential in a competi- tive event (CPE score of 1), thereby making statistical comparisons very difficult, and/or at times, inappropriate. For example, since only two of the 97 Phase III players were assigned a CPE score of 1, the investigator was forced to combine the CPE groups of 1 and 2, and.4 and 5 together for the MANCOVA statistical comparisons. The act of combining different CPE groups together, however, and viewing the new combination as a homogeneous unit may have in part contributed to the violation of homogeneous diSpersion matrices. The low percentage of subjects who were assigned a CPE score of 1 may not have necessarily reflected poor sampling procedures, but rather may have mirrored the entire population of tennis players. 131 Intuitively, it would seem that players who perform below their potential 81-100% of the time would likely get discouraged and drop out of competitive tennis events. Logically, then, the population of tennis players with a CPE rating of 1 might be extremely small. Another CPE scale difficulty resulted as a consequence of group- ing players with CPE ratings of 1 and 2, and,4 and 5 to form the ineffective and effective criterion groups, respectively. When these criterion groups were utilized for various comparative purposes, the responses from players who were assigned a CPE score of 3, were dis- regarded. This middle group, however, was likely representative of the largest portion of ineffective players with whom coaches must contend. That is, these players were defined as playing below their potential in competitive events approximately 50% of the time, and accounted for 37% and 46% of the subjects who completed the STQ-2 and STQ-3, respectively. It seems imperative then, that the responses from these players who have displayed rather serious performance limitations, be included in future data analyses. Thus, before new ways of gathering and/or analyzing data are implemented, it is suggested that the existing CPE scale be revised. The new CPE scale could be arranged along a continuum (see Figure 4) so that the instrument would be more sensitive to individual differences than the existing 5-point scale. Such a design could incorporate the use of a multiple regres- sion analysis, rather than a discriminant analysis, which might make subtle player differences along the CPE continuum more discernable. Furthermore, the inclusion of the present middle group (players who were assigned a CPE score of 3) would be included in the analysis. 132 Please circle the percentage which best exemplifies how often you play up to your potential, against an opponent of the same sex and skill level, in a competitive tennis match. 100 95 90 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0 Figure 4. Proposed Revision of CPE Scale Corollary Research In addition to the revision of the existing CPE scale, research- ers could also implement a variety of corollary investigations. For instance, various within group comparisons of specific CPE groupings could be made with gender, skill, and age as independent variables. Researchers should also assess the cognitive consistency of players' self-talk with respect to changes in situational variables, such as match outcome, Opponents' skill, and match importance. That is, could an athlete's self-talk best be described as a trait or a state phenomenon? Furthermore, it would be beneficial to assess not only the type of self-talk emitted, but also the frequency with which it occurs. Similarly, researchers should assess if the athletes engage in self- talk prior to, during, and/or after skill execution. Audio-video tape procedures might be a useful aid in this type of assessment. Van Noord (1980) in an exploratory, Uncontrolled investigation, found 133 that volleyball players, upon reviewing an audio-video tape of their performance after a match, could easily recall their self-talk before, during, and after the execution of a particular skill or play. The use of audio-video tape recall might also be helpful in assessing the athletes' self-talk during different portions of the match when, for example, they appear to be in a slump and playing inconsistently, as compared to a "hot streak" when they are playing consistently well. Future research efforts parallel to the current investigation should also employ a multidimensional approach in which concurrent anxiety measures, such as Marten's (1977) SCAT, and/or attentional control measures, such as Nideffer's (1978) TAIS could be employed along with the STQ-3 (or revised version). The use of these con- current questionnaires would help to determine the extent of the relationship between athletes' Inappropriate/Appropriate self-talk and corresponding levels of anxiety and attentional skills. In addition to the aforementioned corollary research suggestions, serious considerations should also be given to an experimental manipu- lation of athletes' self-talk. For example, one such study could focus on how the controlled, systematic elimination of an athlete's Inappro- priate self-statements might affect baseline performance patterns. Such experimental manipulations should be thoroughly assessed before self-talk modification programs are advocated for widespread use to enhance athletic performance. While the research directions discussed thus far are not meant to be all inclusive, it is hOped that this investigation, along with 134 the suggestions for future study, will serve as a catalyst for future analyses of competition-related self-talk. APPENDICES 135 APPENDIX A STQ—1 136 SELF-TALK QUESTIONNAIRE (STQ—1) PURPOSE: This questionnaire is designed to see how well you can remember the types of thoughts that cross your mind when you compete in a tennis match. DIRECTIONS: Please respond to the statements on the next two pages according to your SINGLES match play, when recently competing against an opponent of your sex and compara- ble skill level. If you generally DON'T play singles, check here and answer according to your recent doubles match play. For each of the statements on the following pages: FIRST, read the statement describing the competitive situation. SECOND, close your eyes, and try to visualize, as clearly as possible, the last time that you were in that particular competitive setting. Try to remember what the court surface and the surroundings were like. Visualize your opponent. Try to remember how you were feeling (ex. calm, nervous, high spirited, depressed, energetic, shaky, etc.). THIRD, once you have a vivid mental picture of yourself in that situation, try to remember the thoughts that crossed your mind. Generally, the more recent the situation you try to remember, and the more vivid the picture you can re-create in your mind, the easier it will be for you to recall what you were saying to yourself. The following are some typical examples of what tennis players say to themselves while competing: "Look where s/he's standing! I'll hit down the line!‘ "Hit in front." "Rats! I'm playing awful. What are my friends thinking?" "Ace! Alright!" "My second serve is lousy." "Darn! I blew it again." "Yea! Tough hit!" 137 138 FOURTH, write down any recollections of what you were saying or thinking to yourself, as soon as they cross your mind. Don't worry about grammer, com- plete sentences, spelling, or punctuation. Simply write down a word, phrase, or sequence of thoughts as soon as you think of them. PLEASE DO NOT CENSOR YOUR THOUGHTS! Your responses will remain confi- dential. FIFTH, if you did not say anything to yourself during one of the given game situations, or if you can't remember what you said to yourself at that time, WRITE: "NO RECALL" or "NO SELF-TALK" for that particular situation and move on to the next one. Recall and record the things that you remember saying to yourself while experiencing the following game situations, if applicable. 1. You were leading in the set. What, if anything, did you say to yourself between: a. Points b. Games You were behind in the set. What, if anything, did you say to yourself between: a. Points b. Games You were pleased with your: a. Ground Strokes b. Volleys and/or Overheads 139 c. Serves d. Return of Serve You were displeased with your: a. Ground Strokes b. Volleys and/or Overheads c. Serves d. Return of Serve If there is a particular game situation missing, where thoughts frequently cross your mind, please list the situation(s) here and record your corresponding thoughts. APPENDIX B NATIONAL TENNIS RATING PROGRAM (NTRP) 140 A c0py of the National Tennis Rating Program (NTRP) is given below. According to the descriptions given, how would you rate your present skill level? Please circle the appropriate number below. 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 NATIONAL TENNIS RATING PR0GRAM* To Place Yourself A. Begin with 1.0. Read all categories carefully and then decide which one best describes your present ability level. B. Be certain that you qualify on all points of all preceding cate- gories as well as those in the classification you choose. C. When rating yourself, assume you are playing against a player of the same sex and the same ability. NTRP Rating Categories 1.0 This player is just starting to play tennis. 1.5 This player has limited playing experience and is still working primarily on getting the ball over the net; has some knowledge of scoring but is not familiar with basic positions and pro- cedures for singles and doubles play. 2.0 This player may have some lessons but needs on-court experience; has obvious stroke weaknesses but is beginning to feel comfort- able with singles and doubles play. 2.5 This player has more dependable strokes and is learning to judge where the ball is going; has weak court coverage or is often caught out of position, but is starting to keep the ball in play with other players of the same ability. 3.0 This player can place shots with moderate success; can sustain a rally of slow pace but is not comfortable with all strokes; lacks control when trying for power. 3.5 This player has achieved stroke dependability and direction on shots within reach, including forehand and backhand volleys, but still lacks depth and variety; seldom double faults and occasionally forces errors on the serve. *Source: National Tennis Association, 1981. 141 4.0 4.5 5.0 5.5 6.0 6.5 7.0 142 This player has dependable strokes on both forehand and backhand sides; has the ability to use a variety of shots including lobs, overheads, approach shots and volleys; can place the first serve and force some errors; is seldom out of position in a doubles game. This player has begun to master the use of power and spins; has sound footwork; can control depth of shots and is able to move opponent up and back; can hit first serves with power and accuracy and place the second serve; is able to rush net with some success on serve in singles as well as doubles. This player has good shot anticipation; frequently has an out- standing shot or exceptional consistency around which a game may be structured; can regularly hit winners or force errors off of short balls; can successfully execute lobs, drop shots, half volleys, and overhead smashes; has good depth and spin on most second serves. This player can execute all strokes offensively and defensively; can hit dependable shots under pressure; is able to analyze opponents' styles and can employ patterns of play to assure the greatest possibility of winning points; can hit winners or force errors with both first and second serves. Return of serve can be an offensive weapon. This player has mastered all of the above skills; has developed power and/or consistency as a major weapon; can vary strategies and styles of play in a competitive situation. This player typi— cally has had intensive training for national competition at junior or collegiate levels. This player has mastered all of the above skills and is an experienced tournament competitor who regularly travels for competition and whose income may be partially derived from prize winnings. This is a world class player. APPENDIX C COMPETITIVE PSYCHOLOGICAL EFFECTIVENESS SCALE (PHASE I) 143 PERSONAL DATA Male Female Age Where are you presently living? City State Do you play on a tennis team? Yes No If so, what type of team do you play on? High School College Club Other (specify) Approximately how many: a. years have you been playing tennis b. hours of tennis do you play per week The following is a player effectiveness scale. Given your current skill level, it assesses how often you play up to your potential in a competitive situation, and how often you win when playing against an opponent of comparable skill level. Please put a check in the box to the left of the number which best describes your current level of competitive game effectiveness. Percent Game Player Effectiveness Scale Effectiveness 5 = I ALMOST ALWAYS play to the top of my poten- tial; I almost always win when my opponent's 80 skill level is comparable to mine; I'm a very effective player. 100% 4 = I OFTEN play to the top of my potential; I often win when my opponent's skill level is 60 - 79% comparable to mine; I'm effective more often than ineffective. 3 = I SOMETIMES play to the top of my potential; I sometimes win when my opponent's skill 40 _ 597 level is comparable to mine; I'm effective ° about as often as I'm ineffective. 2 = I INFREQUENTLY play to the top of my poten- tial; I infrequently win when my opponent's 2O - 39% skill level is comparable to mine; I'm ineffective more often than effective. 1 - I RARELY play to the tOp of my potential; I rarely win when my opponent's skill level 0 - 19% is comparable to mine; I'm a very ineffec- tive player. 144 APPENDIX D RATER'S CODING PACKET (PHASE II) 145 5. DIRECTIONS Once you have read through all the definitions/classifications, skim over some of the self-statements to get a feeling for the type of statements you will be coding. When you are ready to begin coding: a. Read the statement and decide what type of statement it is. b. Use a pencil and write either an "i," "d," "s," or "p" (please use lowercase letters) next to the corresponding statement number on the answer sheet. If you aren't sure how to rate a particular statement, skip it and come back to it later. It is important that you try to code each statement as accurately as possible. Similar statements, emitted by different people, have been grouped together on the same line, and have been separated by a semicolon. If you have real difficulty in assigning a statement to one of the four categories, circle the number of that statement on the answer sheet. Once you have coded the type of statement, decide whether that statement was emitted by a skilled (s) or unskilled (u) player. a. Write either a "s," or "u" on the answer sheet in the appro- priate column. If you can't decide whether the statement was emitted by a skilled or unskilled player, write an "x" in the appropriate column. Next, decide whether that statement was emitted by a psychologically effective (e) or ineffective(i) player. a. Write either an "e" or "i" on the answer sheet in the appro- priate column. If you can't decide whether the statement was emitted by a psychologically effective or ineffective player, write an "x" under the appropriate heading. If you have any questions, PLEASE call me at 349-6136. Thanks so much for your cooperation and time! In appreciation, you will be given a check for $50 when you complete the rating task. If you are interested in receiving an abstract of the completed study, give me your fall address and I'll be happy to send you a copy. Sincerely, Nancy Van Noord 146 is: 147 CLASSIFICATION OF SELF-STATEMENTS Irrelevant to Good Play Statements focused on aspects irrelevant to doing a good job, including reasons for wanting to win; consequences of losing; regrets about previous actions; wanting the game over with; negative and/or sarcastic remarks to self with no instructional value; wondering about what an opponent, coach, friend, etc. is thinking; rhetorical questions; etc. Distractions in the playing area (who is watching the match; watching players on the next court; etc.) Mind-wandering to off-task thoughts. Difficulty/Ability Statements related to the game, but focused on the easiness or difficulty of winning. Statements focused on the athelete's ability or inability to play well enough--including mental or physical reasons for not playing up to par, self-accusations, self-praise, wishes to do better, attributing skill or lack of skill to luck or other external forces. Anger or joy over something that happened in the game. Assessment of opponent's physical and/or mental abilitites. Words of encouragement (self-motivational statements) to self related to general playing ability. For example, "Come on, you can do it." Self-Exhortation/Evaluation Specific self-directions/advice related to mental and/or physical skill (technique) improvement. For example, directions on what to do, or how to improve his/her current mental functions (to concentrate; to relax; etc.) or how to improve a particular stroke (hit in front; bend your knees; etc.) or movement, position, etc. before, during, and after a point. Self-evaluation/critique of the physical skill and/or mental facets of his/her performance, including approval or criticism of specific moves, skills, mental functions, etc. For example, ”I'm playing too loose;" "I'm hitting late." Problem-Solving/Perceptual Statements related to strategy (game plans) for present or future actions. Assessment of opponent's moves, court position, skill, etc., and then deciding what would be the best course of action to take. For example, "Hit it deep"; "Add more spin"; etc. Preparing self for an upcoming point. Statements regarding where the ball is, where it's coming, where it's going. Statements focused on where to move within a point. Describing own play or moves within a point. 148 PSYCHOLOGICAL EFFECTIVENESS/INEFFECTIVENESS Note: The psychologically effective/ineffective categories focus entirely on the player's psychological capabilities as opposed to the player's physical skills. Therefore, physical skill should 291; be considered when assigning players to either the psychologically effective or ineffective categories. Psychologically Effective A player who is psychologically effective is defined here as one who plays near the top of her/his potential in a variety of competitive situations. That is one who is steady, reliable, plays well under pressure, rarely chokes, and whose skill level in the competitive setting is just as good as her/his skill level in the practice setting. Psychologically Ineffective A player who is psychologically ineffective is defined here as one who infrequently plays near the top of her/his potential in a variety of competitive situations. That is, one who is unsteady, unreliable, doesn't play well under pressure, frequently chokes, and whose skill level in the competitive setting is frequently not as good as her/his skill level in the practice setting. SKILL LEVEL Skilled A player with a 5.0 NTRP score or higher. Unskilled A player with a 3.0 NTRP score or lower. 149 LEADING 1. Stay loose, one more point. 2. Alright! 3. At last, I'm ahead. 4. Keep playing the same way. Make him/her work for it. 5. Relax, take it easy. 6. Keep the ball in play. 7. I must win every point. 8. Okay, that's one less point I have to score. 9. You have to score and win. Don't let him/her beat you. 10. Gosh, I'm winning by a few points! 11. Hit the ball. 12. I can't slack off. I need to be getting as many points as I can. 13. I'm going to win this point/game. 14. Stay cool. Let the other person make the mistakes. 15. All I've got to do is keep it up. Every point I come closer to winning. 16. I hope that I can win this next point/game. 17. Just stay calm and try your best. 18. Just take it easy and you'll make it. 19. I have to get the next point. I'll try not to get too anxious. 20. Come on Babe. 21. Okay, just one more point/game and you've got it. 22. Come on. Win the next point. 23. I should be able to win this one easy. 24. I have to play well. 25. It'll be great if I can win the next set and match. 26. It's about time. 27. I know I can beat her/him. 28. I have to get more points. 29. Relax! Concentrate! You're ahead. Alright! 30. Don't change a thing. Play one point at a time. 31. Hit it deep. 32. I'm ahead. 33. I can afford some high risk shots. 34. Keep your eye on the ball till it hits the racket; watch the ball. 35. Lean into the shot. 36. Play like practice. 37. Hit in front. 38. Keep your wrist firm. 39. Watch the angle of your racket head. 40. Don't think about the score. 41. Play your game. 42. Easy strokes. 43. Play it safe; play percentage tennis. 44. Play her/his weakness. 45. People are watching. I want to look good. 46. Stay on your toes. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 150 LEADING I hope I don't lose. “—- Just concentrate on the ball. Stay calm. I wonder how long I can keep this lead? I've got to keep playing tough; stay tough. Don't get too confident. He/she will come right back. I hit a lucky shot. Things are going my way for a change. The ball went right where I wanted it to go. She/he is tough, but I'm better. This is a boring/great match. Make him/her move. My shots feel great. I hope it lasts. The player on the next court is good/awful. Good! She/he is getting ticked at her/himself. I'm ahead now, but it probably won't last. I wonder how long my luck will hold out. It worked. You're ahead, but that doesn't mean anything. Keep pushing. Yea, that's the way to play. Be ready to serve right away. Keep the momemtum going. Don't go crazy on shots. Play it safe. Stay loose, play each point. Move your feet. Accelerate. Alright! Don't change strategy now. Do the same thing again. I'm ahead. Don't lose your cool. Concentrate on execution. Just 1 (2, 3, 4) points to go. Come on, get tough. Close her/him out. Don't let up. He/she's not as good as you. I finally won a game. Move your feet. You're too relaxed Terrible shot; lousy hit; bad shot. Dummy. Watch the ball. No unnecessary hot dog shots. Get the first serve in. Let's ace this one. Show off. Push. Good recovery. Way to go! The fans think that's great. Good try. Be ready. It's game/set/match point. Now don't blow it like you usually do. Everything is under control. Just do it and enjoy it. Don't beat yourself. Make the other player win the point. Don't give it away. I need one point to win, and I miss an easy shot. Darn, I choked again. You're playing well; I'm looking good; my strokes feel good. Keep it going; keep it up. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 151 LEADING Good play; good job; good shot; tough hit. Way to go, I got her/him now. I'm getting better. Be on guard, he/she can come back. This is my chance. I'll try to make up for my mistakes with this game. Great, but the set's not over. Fantastic. I'm ahead! If I win the lst (2nd, 3rd, etc.) game, the 2nd (3rd, 4th, etc.) one will be easier. Don't mess up now. Alright, won that one. I have to take it one point at a time. It's almost over now. Don't let her/him win many ponts. I can't believe I'm really winning. I'm going to play harder. I'm going to start hitting harder. Chalk this one for me! Alright. Keep it up. Don't let down. I thought that she/he was a better player. I have to keep coming to the net to attack. Watch what you're doing, and do it right, and you'll stay ahead. I'm feeling very up. I can't believe I didn't mess it up. Concentrate! Don't let down. Do better. I've got to keep ahead. Just a few more games. I think I can take this set. Her/his backhand/forehand is weak. He/she doesn't move fast. Play as if the score were close. If you're not careful, you'll lose your lead. She/he's starting to make mistakes. Come on. Play tough. I wonder if I'm going to win this match? Forget everything else. Just concentrate on this game. I'm going to lose this set/match. Concentrate. Play the same as in the other games I won. Keep the pressure on. Get the ball back deep and high. Keep it up, you've got the momentum. Clear your mind of past points. You've got her/him. Stay on top. Don't let up. Use your shots. Come on, you can do it. Keep playing steady. O‘ U-L‘LAJNt—i \OCDN 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 152 BEHIND Come on, loosen up. Gotta have this one. Keep it in play. Darn, I blew it again. It's only a few points, don't let down. Press a little but, but don't make mistakes. Make him/her hit winners. Don't worry, you can come back; I'm behind now, but I know I can do it. I've got to concentrate on hitting the ball right. I've got to do my best. I'll try to even up the score and pretend that we're just starting the game. I don't care anymore. I just want to get this match over with. Come on! Be consistent! Come on! You'll make it up the next point. Don't get nervous. Be confident and you'll score some points. Oh Darn! Please let me win just this once. I'm going to lose this set/match. If I lose this game, I'm going to have a tougher time. Darn; Shoot; Rats; etc. If I could get a few points and win a couple of games, I'd be alright. Think 0fthis match as fun, not competition, then I'll do alright. I have to catch up to him/her. Stay calm. I'm getting tense/nervous/uptight, etc. Be more careful on how you hit the ball next time. I have to keep calm because if I try too hard, I'll probably make mistakes. Come on. Concentrate. Watch the ball. Follow through. I suppose you are going to give her/him the next point too. Hit the ball. Relax. Don't swing like a baby. I can't win if I keep playing like this. Take your time. Hit the ball correctly. Don't rush your strokes. Concentrate! Get back in the game. The game is not over. I better get this one to stay in the game. I've got to return that serve. I have to move faster. His/her placement is tough. Lean into the shot. Hit in front. Keep your wrist firm. Watch the angle of the racket head Play our game. Forget about the score. Watch the spin. Your form looks terrible today. PeOple are watching me. I wonder what they're thinking. Stay on your toes. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 153 BEHIND Be aggressive, not cautious. Don't tighten up. I wonder how my teammates are doing. Stupid; Fool; Idiot; Dummy; etc. Bend your knees. Come on, you can do better than that. Don't hit so hard. Oh no, here I go again. What is my problem? I hope no one I know is watching me. I can't lose the next point/game. I'm no good. Sometimes my game stinks. I'm not feeling well. I'll be glad when this match is over. My muscles are sore. I've got to start pushing. Make her/him move. I hate to lose. I never win. I'm choking; I'm blowing it. Stay in the point as long as possible. Push! No mistakes! Get going; Let's go; Got for it; Go to it. Be more aggressive. Keep the ball to his/her backhand. I'm losing. Just try to keep the ball in play. Don't panic. Come on, just 1 (2, 3, 4) points. I'm hitting late. Move your feet. Relax. Hit out. You are playing the ball. Get your act together. Get tough. He/she's not as good as you. Get this game. Why did I hit that shot? Terrible shot; Lousy return; Bad shot. Missing your serve too much. Get warmed up. Keep the ball deep. Stay in the point. Relax. Get more balls in play. Settle down, I need this one. You can come back. Start hitting the ball the way you should. It's only this game. This next game, I'd better do better. This is going to be my game. My stupid opponent thinks she/he is so good just because she/he is winning. I'm getting angry. Oh no! What on earth am I doing? O.K., take one point at a time. You'll catch up. Come on. Gotta work harder on those shots. There is still hope, so don't give up. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. 154 BEHIND He/she is a lot better than me. Got to get ahead. I'm doing terrible. Maybe next time. Oh brother! Why am I playing so bad? I know that I can play better. The harder I try, the more I mess up. I have to win this game so that I can catch up. It's just a game. There is no way I will catch up. I've got to think more clearly. Self! Whose side are you on? Better get started. Don't get too far behind. Play tennis. Is she/he that good, or am I that lousy? I have to come back and play well to win the match. Think about what you're doing. I should just forefit. It's alright. This game is mine. Be more aggressive. Out smart him/her. Lucky break, I'll come back. Let her/him make the mistakes. Come on, play our game. Play like practice. Don't give the game away. Make him/her work for it. My game is off; my strokes are off. I'm getting tired; I'm exhausted. Play it safe. Go for the high percentage shots. Get behind the ball. Forget playing cautious. It's not working. Be aggressive. You can't do any worse. I'm going to lose anyway, so go for it. I wish that I could look like that player on the next court. I don't like this court surface. The heat/cold is starting to get to me. Concentrate! Forget about everything else, just concentrate on the game/the ball. Okay, I made a mistake. Forget it. Concentrate on the next point. Don't make the same mistake twice. Way to give it away. Every time I try to be cautious, I lose the point. Play tennis much? Take your time. Slow everything down. Don't think so much. I'm trying too hard. Try something new. Find a weakness; Change the game. Make something happen. How can I change my strategy to win? I need a break. 139. 140. 141. 142. 143. 144. 145. 146. 147. 148. 149. 150. 151. 152. 153. 154. 155. 156. 157. 158. 159. 160. 161. 162. 163. 164. 165. 155 BEHIND I need to hold serve. "““‘ If I lose, I don't want to lose by a big margin. Get busy. What do you have to lose? Stay with it. Play her/his faults. Let's look at what's happening here. Just play a stall type of game, not too fancy. Not too many outright winners, but when opportunity knocks, be ready. You've got to get tough. Serve that ball in. Don't be so chicken to use your shots. Play steadier. Push harder. Wasn't concentrating on the shot, but forget about it. Move! You're too slow. You'd better get the next one. I'm still thinking about my missed shot. Let's ace this one. Just get it over. Good recovery. How can I get into this game? How can I fire myself up? I feel crummy. I'm cold and not ready to play. I should have gotten that. Why did I just stand there? Go for it. That was awful. Good try. I wonder what my Dad/Mom/Coach is thinking? Be ready. Come on. Relax. You can do it. That ball was in! What a cheater! Maybe I'll get lucky and get it together. HHHHkOCDNO‘Ln-waI-n WND-‘OO o o o o O .0. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 26. 27. 28. 29. 30. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 156 PLEASED WITH PLAY Good shot; Tough hit; Way to hit; Nice hit; Good serve. Felt good. Way to move. Good; Alright; Okay; Fantastic; Yes; Yea; Great; Nice. My practice is paying off. Alright! Gotta hit those kind of shots more often. Good job, everyone noticed my strength. I'm surprised I made that shot. Hey! That was a good one. That's nice and smooth. Gee, beauty in motion, aren't I? My backhand is the best. My forehand is good. I'm hitting nice and deep with a lot of topspin. I got it! Not bad! I haven't messed up that bad. I'm really doing well today. Yes, It just barely went over. Good powerful shot; Good speed and depth; Nice and solid. I'm watching the ball. I'm hitting in the middle of the racket. I'm getting in good position. Keep your wrist firm. Smooth follow through. Good placement. Good, she/he can't reach it. I hope that I can keep this up. I don't know what I'm doing right, but I hOpe it lasts. Hit short; Deep; Down the line; Cross-court; to the corner, etc. He/she can't handle my hard forehands/backhands. Bend your knees; stay low. Try to enjoy yourself. Oh no, a short shot. Alright, I got it! Good! I hope I don't hit this into the net/out of bounds. I can't lose this point/game. Move your feet. Hit . . . N33. Bounce . . . Hi2. Lucky shot. I'll be glad when this match is over. This player should be easy to beat. This is a great/awful match. Keep pushing. Make him/her work for it; Make him/her move. Ace her/him; Kill it; Put it away. Stay tough; Stay solid. Go for it. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 157 PLEASED WITH PLAY I wonder how long my luck will last. It worked! Alright, a lob, kill it! I love these kinds of shots. Move up/back. Yea! That's the way to play; That's it! Hit to where he/she is running from. Get her/him out of position; Set her/him up; Keep her/him off balance. Come on, land in. Everything is going well. Don't think so much. Keep your mind on the ball. Watch the ball; Watch the seams on the ball; Keep your eyes on the ball for as long as possible. Stay loose. Don't rush the passing shot. Good job. Keep it going; Alright! Keep it up. Early preparation. Stay in the zone. Don't get overejoyed. Step it to the volley. Hit up. Move forward at every opportunity. I'm proud of myself. Right on the baseline. What a relief. I'm improving. Wow, did I do that? I just can't believe it! I'm giving it everything I have. Good shot, but it was an accident. Remember what you did right so that you can do it again. Yes, but I could still improve on it. Fair shot. Good shot, but it was an accident. I got lucky. Thank goodness, once in a while I can do it right. Wow! I'm a wall. Gee, what can I say? Maybe today is my day. My overheads/volleys/strokes/serves are good. I'm shocked, I usually foul those up. Keep the ball in front; Hit in front. Step into the shot; Go toward the ball; Forward. Ah, perfect height. It's too high, back up. I can reach it. Don't kill it. What if I miss it? Don't be fancy. This is my point/game/set/match. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 158 PLEASED WITH PLAY Keep a good grip. Keep closing. Kill it. No change. Keep it up. It feels good. Why can't I do this all the time. I know I can do it. Then why can't I? Keep it low and deep. Execute well. Don't get reckless. Don't get too introspective. Leave it alone. Keep your mind on the upcoming events. Keep attacking; be aggressive. Keep the toss high. Hit it on the line . . . good! I wish I could hit like this all the time . . . especially when I need to. She/he must really be a good player to return that serve. I'll bet he/she can't hit this serve. Hey, I should keep serving this way. This is the way I want to serve, to scare her/him. I'm going to win. Good. Now you're extending. Keep it up. That's more like it. Now a little more stuff on the first serve. Very effective. I know I can do it. I knew I could serve well. I finally kept my eye on the ball. Reach! Concentrate! I almost never double fault. I'm dragging my toe again. My shoe is going to fall apart. Don't think, just hit. Get on t0p of the ball; Hit on the rise. I need this point/game. Don't double fault. Serve tough. Wider; play the angles. Harder; keep her/him deep. Okay. Mix it up. What exactly did I do? Play her/his weakness. 3/4 pace with spin. Don't over serve. Good serving. Keep the 1st serve going in. Swing, don't block. Try down the line, keep him/her awake. Just keep on the ball. I love it! Do it again. I received that well. That was a hard hit. Wow! She/he could have aced me, but I got it! When I get too confident I miss the next point. Finally, you're doing something right; It's about time. 143. 144. 145. 146. 147. 148. 149. 150. 151. 152. 153. 154. 155. 156. 157. 158. 159. 160. 161. 162. 163. 164. 165. 166. 167. 168. 169. 170. 171. 172. 173. 174. 175. 159 PLEASED WITH PLAY Take that! Watching the ball all the way pays off. I'm on and off. Just get it back. I'm too excited. Those aren't that hard. I'm worried about his/her serve. I'd better be careful. Ready . . . good return. Look where she/he is standing. Be ready to move. Oh good, it's a forehand/backhand. Watch the spin. Attack on her/his second serve. Play it safe. Don't blow it. It feels good but scary. I don't know if I can keep it up. Remember they are just ground strokes. Try to go in as soon as possible. Good, put the pressure on him/her. Stay alert. Keep feet moving. Watch her/his toes. Show off. Good recovery. I'd better hit this good and hard. Nice try. Good. That should please my coach. Alright! A wide open court. Don't blow it. I got it, hit around him/her. Get into position. Anticipate! Oh no, my shot is going out. Alright, the sucker hit it. Lucky me. Everything is under control. Just do it and enjoy it. \OCDNO‘kfl-L‘UJNH o o oo o 11. 12. 14. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 160 DISPLEASED WITH PLAY Don't Open the face. I'm no good. Keep the ball in play; Get the ball back. Come on! Hit it right. Turn your racket a little more. Is this the best I can do? Not high enough. Stupid me. There's one you should have scored on. Better make up for it. Poor shot; Terrible shot; Lousy hit; Bad shot; etc. Lousy. What is my opponent thinking? Needs improvement. Darn; Shoot; For heaven sakes; Golly; etc. I've got to try and do it better, and not waste my time getting angry. Oh, come on! I think the net is too high, that's why it didn't go over. What am I doing wrong. Let's get it together here. Why don't you just leave the court while you have some dignity left. Back to the drawing board. Idiot; Stupid' Fool; Dummy; Jerk; etc. Watch the ball not the opponent; not the area you're hitting to. Stroke the ball. My forehandsi are too high. My thoughts aren't on the game; I'm not concentrating. I'm not mentally ready for this game. Shoot! Do better next time; Come on, you can do it; You can do better than that. Calm down. You'll have another opportunity. Next time. Forget that one. I need more practice. Get in better position. Just one point at a time. Hang tough; Be tough. Bend you knees. Play like practice. Don't think/worry about the score. I hope I don't hit this into the net/out of bounds. I wish those people would st0p watching me. Concentrate on your form. If you're going to make mistakes, you might as well look good. Play safe. Return to the middle. My strokes stink today; I'm playing terrible; I can't hit worth a darn today. Just concentrate on the ball; watch the ball. Hit . . . N23. Watch the seams on the ball; Watch the spin. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 161 I'm not feeling well. DISPLEASED WITH PLAY My muscles are sore. I've got to start pushing. Make her/him work for it. Don't give it away. Don't give up. Start over again. Forget about everything. Just concentrate on the next point. Okay, I made a mistake . . . forget it. Don't make the same mistake twice. Try hitting the ball for once. I'm exhausted. Racket back early. Set up. Don't try so hard. Slow up. No mistakes. Don't rush the shot. Hit/aim higher; Lift more; Place the ball up; Hit up. Slow down the stroke. Don't hit so hard. I feel foolish. Why don't you concentrate? Find the timing spot. Go after it. Don't be timid. Punch, don't swing on the volley. Maybe next time. Come on. Think about it. Rats. I hit it too hard. This is not my day. I never get those anyway. If I could only be a few inches taller. Darn! I need to work on my overheads/volleys. I knew I shouldn't have done that. Why did I even try; Why did I hit that shot? Darn. I'm just no good at the net. It's not the racket. Don't rationalize. Play the game. Should have let the ball bounce. Cut out the fancy shots. Hit harder. Hit deeper. Come on, stOp messing around. Check things out. Take it easy. Relax. Concentrate. I wasn't ready. Don't choke. Typical! Here I go again. Calm down. Move faster; Move your feet; Stay on your toes. Step into the shot; Lean into the shot. Hit in front; Meet the ball; Hit it early. Oh well, at least your form looked good. Keep your wrist firm. I hOpe I don't hit this out/into the net. Come on, win this point. I'm choking. Hang on to the racket. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 162 DISPLEASED WITH PLAY Way to give it away. Oh brother, right to his/her forehand. I'm going to get killed. Play tennis much? Stay solid/low. Stay back. Keep the pressure on. You're dropping your head. Watch the ball, Stupid. Concentrate! Come on it late. Shorten your backswing. You're overplaying the ball. Step into the volley. I keep spraying balls. Get warmed up. Reach for the ball. Control the shots more. Gotta have a good toss. No double fault. Okay, gotta have this one. I'm going to have to serve better or make it up in other ways. I must be getting tired. Stop cutting your strokes. I must control myself. What happened to those good serves I served during practice. My serve is weak. If I could just have a good serve, I could win more. Darn it! You never double fault. Why now? My opponent must think I'm a real loser. Come on, get your act together. StOp. Take a deep breath. Try a little harder. Why can't I ever get them in? You have to do better. Look out . . . here it comes. Hit with confidence. Be deliberate. What a screw up. There's no reason for that! I made a careless mistake. Adjust the toss. Lousy toss . . . Why did I hit it? My serves are off. Watch the angle of the racket head on contact. Ace her/him. Hit down the middle; To his/her backhand/forehand. Pull her/him out of position. Mix it up. Throw him/her off. I hOpe I don't double fault. What is my problem; What is wrong with me? I'm going to lose. I wonder why I like this sport? I'll be glad when this game/set/match is over. I never win. I'm choking again. Concentrate. Don't think so much. Just let it happen. How embarrassing. I hope no one is watching me. 143. 144. 145. 146. 147. 148. 149. 150. 151. 152. 153. 154. 155. 156. 157. 158. 159. 160. 161. 162. 163. 164. 165. 166. 167. 168. 169. 170. 171. 172. 173. 174. 175. 176. 177. 178. 179. 180. 181. 182. 183. 184. 185. 186. 187. 163 DISPLEASED WITH PLAY I keep losing the ad point. I'm tightening up. Relax. Take your time. Slow the game down; Don't rush it. Get it in play. More spin. You're playing too loose. 3/4 pace with depth. I'm hitting too hard. Keep your eye on the ball longer. Weight on back foot longer. No serve with this new racket. Missing your serve too much. Oh no, it's a short shot. Darn, I was off balance. She/he must have loved that . . . burning me. Why don't you think? That was dumb. Be ready. You can get this one. Watch the ball. Get your head in the game and get the next return. I have to concentrate more and not get so tense. It's okay. Be alerttx>make a better return. I should have been able to hit that/ How could I miss that shot? I'd better shape up. Now my opponent is going to think that he/she is hot stuff. Not again. Watch the ball. Don't blow it. Might as well not play if you can't return the serve. I don't follow through. I need work on my second serves. You weren't in the right place. Darn it! Pay attention. Something has to be wrong with me. It's no big deal, just concentrate. Improve. Darn it all! Why is he/she doing this to me? Hit the ball in the middle of the racket. Come on. The serves aren't that tough. Try to look like a pro. I feel like a real clutz today. Go for it. You can get it. Late! What was that? Concentrate. You're catching it late. Let's get it earlier. Stand back farther for more reaction time. Watch the timing of ball over the net. Racket in full follow through position. 164 Division: Leading Behind Pleased Displeased i=Irrelevant/ S=Skilled E=Psychologically Unrelated Effective _ U=Unskilled d—ggii:EUIty/ I=Psychologically y X=Undecided Ineffective s=Self— Exhortation/ X=Undecided Evaluation p=Problem solving/ Perceptual 10. 11. 12. 13. 14. 15. 16. 17. 18. APPENDIX E STQ-2 165 PURPOSE: PARTICIPANTS: DIRECTIONS: SELF-TALK QUESTIONNAIRE This questionnaire is designed to see how well you can remember the types of thoughts that cross your mind when you compete in a SINGLES tennis match. Hundreds of tennis players from Florida, Texas, and Michigan with skill levels ranging from intermediate to world class caliber are participating in this inves- tigation. Four competitive SINGLES game situations are listed on the following pages: 1. LEADING in a set 2. BEHIND in a set 3. PLEASED with play 4. DISPLEASED with play Listed under each game situation are examples of common thoughts that have crossed tennis player's minds during match play. For each of the four game situations: FIRST, try to remember the last time that you were in that particular game situation, competing against an Opponent of your sex and comparable skill level. SECOND, close your eyes and try to visualize yourself as clearly as possible while playing under that particular condition. (For example, the last time you were LEADING in a set.) THIRD, once you can vividly recall that portion of the match, open your eyes and read each one of the statements. Rate EACH statement from 1 to 5 according to the scale given below, to indicate how frequently, if at all, you have said that statement (or something closely related to that statement) to yourself while competing under the different game situations. Circle the most appropriate number (circle one number only) imme- diately to the right of each corresponding state- ment. - 166 167 Note: At times similar statements are grouped together on the same line, and are separated by a semicolon. Underline the statement(s) you identify with the most and rate it accordingly. RATING SCALE 1 = NOT AT ALL 2 = INFREQUENTLY 3 B SOMETIMES 4 = OFTEN 5 = ALMOST ALWAYS Your responses will remain strictly confidential. Thank you for your time and cooperation! 168 L E A D I N G You had the Leading score in the set. How often, if at all, did you say the follow- ing statements to yourself? 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. I know I can beat her/him. I hit a lucky shot. Things are going my way for a change. He/she is tough, but I'm better. I can't believe I didn't mess it up. Concentrate on execution. I wonder how long my luck will hold out? I need one more point to win, and I miss an easy shot. Darn, I choked. Play as if the score were close. Don't go crazy on shots, play it safe; play percentage tennis. I'm ahead now, but it probably won't last. I hope I don't lose. I have to play well. Keep it up. Don't let down. You've got the momentum; you've got him/her. Stay on top. Don't let up. Play her/his weakness. I hope that I can win this next point/game. Don't mess up now. It's game/set/match point. Now don't blow it like you usually do. His/her backhand/forehand is weak. Play your game. Stay loose, play each point. I can't believe IHnreally winning. close her/him out . I have to keep coming to the net to attack. I wonder if I'm going to win this match? Don't beat yourself. Make him/her win the point. Don't give it away. .. e As a egg :3 QHEEE’ BREE-'2 eases 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 1 2 3 1 2 3 4 5 1 2 3 4 5 l 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 1 2 3 1 2 3 4 5 1 2 3 4 5 l 2 3 4 5 l 2 3 4 5 1 2 3 4 5 l 2 3 4 5 169 B E H I N D You were Behind in the set. How often, if at all, did you say the following statements to yourself? \OCDNO‘UIbUJNt—I 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. Stupid; Fool; Idiot; Dummy; etc. Darn, I blew it again. There is no way I will catch up. I need a break. What is my problem? Relax. Hit out. Let's look at what's happening here. Why did I hit that shot? Concentrate! Get back in the game. The game is not over. Stay with it. Play her/his faults. I have to come back and play well to win the match. Not too many outright winners, but when Opportunity knocks, be ready. People are watching me. I wonder what they're thinking? The harder I try, the more I mess up. I never win. I need to hold serve. I'll be glad when this match is over. Try something new. Find a weakness; Change the game. Make something happen. He/she is a lot better than I am. Oh Darn! Please let me win just this once. I've got to start pushing. I hope no one I know is watching me. Be aggressive, not cautious. Maybe I'll get lucky and get it together. It's only a few points, don't let down. Oh brother! Why am I playing so bad? I know that I can play better. Press a little bit, but don't make mistakes. I suppose you are going to give her/him the next point too. Get going; Let's go; Go for it; Go to it. Stay cool. Let him/her make the mistakes. Keep the ball deep. Stay in the point. Stay in the point as long as possible. Push! I'm choking; I'm blowing it. I'm no good. >« a as s HEW é ~ 33 .153” E." Egg 2.3 '23” "' mffifig SE53": 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 l 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 l 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 l 2 3 4 5 171 D I S P L E A S E D You were Displeased with your play. How often, if at all, did you say the following statements to yourself? 1. Stay solid/low. 2. She/he must have loved that . . . burning me. 3. Take your time. Slow the game down; don't rush it. 4. If I could just have a good serve, I could win more. 5. You're catching it late. 6. Lousy. What is my opponent thinking? 7. Adjust the toss. 8. What a screw up. There's no reason for that! I made a careless mistake. 9. Hit short; deep; down-the-line; cross-court; to the corner; etc. 10. Shorten your backswing. You're overplaying the ball. 11. Typical! Here I go again. Calm down. 12. My opponent must think I'm a real loser. 13. Hit/aim higher; lift more; place the ball up; hit up. 14. Racket in full follow through position. 15. I hope I don't hit this out/into the net. 16. Watch the angle of the racket head. 17. Oh no, it's a short shot. 18. Next time. Forget that one. 19. Hit with confidence. Be deliberate. 20. I feel foolish. Why don't you concentrate? 21. You're playing too loose. 22. Find the timing spot. 23. Just one point at a time. 24. I hope I don't double fault. 25. Oh brother, right to his/her forehand. I'm going to get killed. 26. Step into the volley. 27. Okay, I made a mistake. Forget it. Concentrate on the next point. 28. This is not my day. 29. Stay back. 30. Watch the seams on the ball; watch the spin. 31. 3/4 pace with depth. I'm hitting too hard. 32. Move faster; Move your feet; Stay on your toes. z» a New at egg :3 2°: 5 Ease Emit-15 OZOFH ZHUJO< 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 APPENDIX F COMPETITIVE PSYCHOLOGIAL EFFECTIVENESS SCALE: PLAYER'S FORM FOR PHASES II AND III 172 PERSONAL DATA Name Age Male Female Name of your tennis club Location: City State Approximately how many: a. YEARS have you been playing tennis b. HOURS of tennis do you play per week The following is a player effectiveness scale. IT HAS NOTHING TO DO WITH YOUR PHYSICAL SKILL. Rather, it assesses your PSYCHOLOGICAL effectiveness. That is, a psychologically effective player is one who is steady, consistent, reliable, plays well under pressure, rarely chokes, and who plays as well(or better) in the competitive setting as in the practice setting. Using the scale given below, please put a check in the box to the left of the number which best describes your current level of COMPETITIVE PSYCHOLOGICAL EFFECTIVENESS. REMEMBER, IT HAS NOTHING TO DO WITH YOU PHYSICAL SKILL. PERCENT MATCH PLAYER EFFECTIVENESS SCALE EFFECTIVENESS 5 = I ALMOST ALWAYS play to the top of my potential; I almost always win when my opponent's skill level is comparable to mine; I rarely choke; I am very steady and consistent. 80 - 100% 4 = I OFTEN play to the tOp of my potential; I often win when my opponent's skill level is comparable to mine; I choke now 60 - 79% and then, but I'm usually steady and consistent. 3 = I SOMETIMES play to the tOp of my potential; I sometimes win when my opponent's skill 40 _ 597 level is comparable to mine; I choke about ° as often as I am steady and consistent. 2 = I INFREQUENTLY play to the top of my poten- tial; I infrequently win when my opponent's 20 skill level is comparable to mine; I choke more often than I'm steady and consistent. 39% 1 = I RARELY play to the top of my potential; I rarely win when my Opponent's skill level is O - 19% comparable to mine; I almost always choke. 173 APPENDIX G COMPETITIVE PSYCHOLOGICAL EFFECTIVENESS SCALE: COACH'S FORM FOR PHASES IT AND III 174 COACH'S ASSESSMENT OF PLAYER RATINGS Your name Name of club Please list the names of your players participating in this study on the last page. In addition, please rate each player according to the two scales given below. Record your responses on the last page. 1. SKILL RATING. A copy of the National Tennis Rating Program (NTRP) is given on the next pages. Please rate each of your players by assigning the number which best describes her/his current level of play. PSYCHOLOGICAL EFFECTIVENESS RATING. A player who is psychologi- cally effective is defined here as one who plays near the top of his/her potential in a variety of competitive situations. That is, one who is steady, reliable, consistent, plays well under pressure, rarely chokes, and one who plays as well (or better) in the competitive setting as in the practice setting. Using the scale given below, please rate each of your players by assigning the number which .best describes her/his current level of psychological effectiveness in the competitive setting. THIS RATING HAS NOTHING TO DO WITH YOUR PLAYER'S PHYSICAL SKILL! PLAYER EFFECTIVENESS SCALE PERCENT MATCH COMPETITIVE PSYCHOLOGICAL EFFECTIVENESS RATING EFFECTIVENESS ALMOST ALWAYS plays to the top of his/her poten- tial; almost always wins when opponent's skill level is comparable to his/hers; rarely chokes; very steady and consistent player. 80 - 100% 4 = OFTEN plays to the tOp of her/his potential; often wins when opponent's skill level is 60 _ 797 comparable to hers/his; chokes now and then, ° but is usually steady and consistent. 3 SOMETIMES plays to the top of his/her poten- tial; sometimes wins when opponent's skill level is comparable to his/hers; chokes 40 - 59% about as often as he/she is steady and consistent. INFREQUENTLY plays to the top of her/his poten— tial; infrequently wins when Opponent's skill 20 level is comparable to hers/his; chokes more often than she/he is steady and consistent. = RARELY plays to the tOp of his/her potential; rarely wins when opponent's skill level is O - 19% comparable to his/hers; almost always chokes. 175 176 PLAYER' NAME SEX SKILL RATING PSYCHOLOGICAL EFFECTIVENESS RATING 10. 11. 12. 13. 14. 15. l6. 17. 18. 19. 20. 21. 22. APPENDIX H DISCRIMINANT FUNCTION ANALYSIS ON 118 STQ-2 STATEMENTS 177 Table H. Results of Discriminant Function Analysis on 118 STQ-2 Items Item3 Wilks' Lambda F Significance L1 .97428 2.35 .1289 L2 .98874 1.01 .3167 L3 .97044 2.71 .1032 L4 .98879 1.01 .3179 L5 .97236 2.53 .1153 L6 .95526 4.17 .0441 L7 .79691 22.68 .0000 L8 .99827 .15 .6954 L9 .98222 1.61 .2077 L10 .92230 7.50 .0075 L11 .98551 1.31 .2558 L12 .96295 3.42 .0676 L13 .98791 1.09 .2995 L14 .98969 .93 .3382 L15 .99024 .88 .3514 L16 .92509 7.21 .0087 L17 .94779 4.90 .0294 L18 .98870 1.02 .3159 L19 .95987 3.72 .0569 L20 .95810 3.89 .0516 L21 .8777] 12.40 .0007 L22 .94887 4.80 .0311 L23 .99997 .25 .9606 L24 .99435 .51 .4788 L25 .99209 .71 .4017 B1 .96629 3.11 .0815 B2 .92033 7.71 .0067 B3 .97688 2.11 .1502 B4 .96822 2.92 .0909 BS .84360 16.50 .0001 B6 .99141 .77 .3823 B7 .99901 .88 .7674 B8 .98896 .99 .3216 B9 .99883 .10 .7473 B10 .99371 .56 .4548 B11 .92140 7.59 .0071 B12 .98103 1.72 .1929 813 .96701 3.04 .0849 B14 .98629 1.24 .2691 B15 .89038 10.96 .0013 816 .98686 1.19 .2792 817 .89771 10.14 .0020 B18 .99999 .82 .9772 B19 .94673 5.01 ' .0277 (table continues) 178 179 Table H. (Continued) Itema Wilks' Lambda F Significance BZO .93533 6.15 .0150 B21 .99616 .34 .5596 822 .82562 18.80 .0000 323 .99935 .58 .8101 B24 .89406 10.55 .0016 325 .99958 .38 .8467 B26 .93393 6.30 .0139 B27 .99985 .14 .9070 328 .75369 29.09 .0000 829 .98792 1.09 .2996 B30‘ .99929 .63 .8017 B31 .99995 .42 .9488 B32 .99801 .18 .6743 B33 .81312 20.46 .0000 B34 .73240 32.52 .0000 P1 .98309 1.53 .2193 P2 .98463 1.39 .2416 P3 .98712 1.16 .2841 P4 .93548 6.14 .0151 P5 .93610 6.08 .0156 P6 .99873 .11 .7373 P7 .97987 1.83 .1798 P8 .99624 .34 .5639 P9 .99950 .45 .8328 P10 .98549 1.31 .2555 P11 .99318 .61 .4364 P12 .99983 .15 .9032 P13 .99660 .30 .5830 P14 .99732 .24 .6263 P15 .96887 2.86 .0943 P16 .99147 .77 .3839 P17 .99403 .53 .4666 P18 .98047 1.77 .1865 P19 .99924 .68 .7951 P20 .99720 .25 .6184 P21 .99975 .22 .8821 P22 .96251 3.47 .0659 P23 .99927 .65 .7996 P24 .99868 .12 .7324 P25 .99767 .21 .6492 P26 .99846 .14 .7115 ,P27 .99999 .67 .9794 D1 .9881] 1.07 .3035 D2 .98893 1.00 .3210 D3 .99139 .77 .3817 D4 .98991 .91 .3434 (table continues) 180 Table H. (Continued) Itema Wilks' Lambda F Significance D5 .93931 5.75 .0186 D6 .97719 2.08 .1530 D7 .99160 .75 .3874 D8 .90006 9.88 .0023 D9 .99933 .59 .8081 D10 .99823 .16 .6919 D11 .88183 11.93 .0008 D12 .93880 5.80 .0181 D13 .99889 .98 .7544 D14 .97731 2.07 .1541 D15 .98063 1.76 .1883 D16 .99982 .16 .8990 D17 .94918 4.77 .0317 D18 .99952 .42 .8372 D19 .99954 .41 .8409 D20 .90344 9.51 .0027 D21 .99859 .13 .7240 D22 .99763 .21 .6464 D23 .99986 .13 .9109 D24 .92417 7.30 .0082 D25 .97106 2.65 .1070 D26 .99803 .18 .6759 D27 .96541 3.19 .0776 D28 .84709 16.07 .0001 D29 .99964 .32 .8579 D30 .99139 .77 .3816 D31 .96019 3.69 .0579 D32 .99964 .32 .8589 3The letter which precedes the STQ-3 item number indicates the game situation category to which the statement belongs: L = Leading in a set B = Behind in a set P = Pleased with play D = Displeased with play APPENDIX I STQ-3 181 PURPOSE: PARTICIPANTS: DIRECTIONS: SELF-TALK QUESTIONNAIRE This questionnaire is designed to see how well you can remember the types of thoughts that cross your mind when you compete in a SINGLES tennis match. Hundreds of tennis players from Florida, Texas, and Michigan with skill levels ranging from intermediate to world class caliber are participating in this investigation. Four competitive SINGLES game situations are listed on the following pages. 1. LEADING in a set 2. BEHIND in a set 3. PLEASED with play 4. DISPLEASED with play Listed under each game situation are examples of common thoughts that have crossed tennis players' minds during match play. For each of the four game situations: FIRST, try to remember the last time thatsmn1were in that particular game situation, competing against an opponent of your sex and comparable skill level. SECOND, close your eyes and try to visualize yourself as clearly as possible while playing under that particular condition. (For example, the last time you were LEADING in a set.) THIRD, once you can vividly recall that portion Of the match open your eyes and read each one of the statements. Rate EACH statement from 1 to 5 according to the scale given below, to indicate how frequently, if at all, you have said that statement (or something closely related to that statement) to yourself while competing under the different game situations. Circle the most appropriate number (circle one number only) immediately to the right of each corresponding statement. NOTE: At times similar statements are grouped together on the same line, and are separated by a semi— colon. Underline the statement(s) you identify with the most and rate it accordingly. RATING SCALE 1 = NOT AT ALL 4 = OFTEN 2 = INFREQUENTLY 3 = SOMETIMES 5 ALMOST ALWAYS Your responses will remain strictly confidential. Thank you for your time and cooperation! 182 183 L E A D I N G You had the LEADING score in the set. How often, if at all, did you say the following statements to yourself? 1. He/she is tough, but I'm better. 2. I wonder how long my luck will hold out? 3. I need one more point to win, and I miss an easy shot. Darn, I choked. 4. Concentrate on execution. 5. I'm ahead now, but it probably won't last. 6. Play your game. 7. I hope I don't lose. 8. I hOpe that I can win this next point/game. 9. Don't mess up now. 10. It's game/set/match point. Now don't blow it like you usually do. 11. Stay loose, play each point. 12. Close her/him out. 13. I can't believe I'm really winning. .. 2 saw 3 egg :3 2'38“ 5; .335. BREE-Ia OZOFI-c szo< 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 184 B E H I N D You were BEHIND in the set. How often, if at all, did you say the following statements to yourself? 10. 11. 12. 13. 14. 15. 16. 17. 18. Stupid; Fool; Idiot; Dummy; etc. Darn, I blew it again. What is my problem? I have to come back and play well to win the match. People are watching me, I wonder what they're thinking? Not too many outright winners, but when Opportunity knocks, be ready. The harder I try, the more I mess up. I never win. I'll be glad when this match is over. He/she is a lot better than I. Oh darn! Please let me win just this once. I hOpe no one I know is watching me. Maybe I'll get lucky and get it together. Oh brother! Why am I playing so bad? I know that I can play better. I suppose you are going to give her/him the next point too? Get going; Let's go; Go for it; Go to it. I'm choking; I'm blowing it. I'm no good. .. a .33 g 4E0: d we as: :3 Hggeo O 0E5 ZHUJO<2 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 I3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 185 P L E A S E D You were PLEASED with your play. How often, if at all, did you say the following statements to yourself? Early preparation. Finally, you're doing something right; It's about time. I hope that I can keep this up. It feels good but scary. I don't know if I can keep it up. I'm surprised I made that shot. Good, put the pressure on her/him. It feels good. Why can't I do this all the time? Watching the ball all the way pays off. Alright! Awide open court. Don't blow it. >3 g 3130) 335 35% s 38': '7 a: n: 53 Z; Buzz-us: ozong zv-umo 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 186 D I S P L E A S E D You were DISPLEASED with your play. How often, if at all, did you say the following statements to yourself? 10. 11. What a screw up. There's no reason for that! I made a careless mistake. You're catching it late. Okay, I made a mistake. Forget it. Concentrate on the next point. Typical! Here I go again. Calm down. My opponent must think I'm a real loser. I hope I don't hit this out/into the net. 3/4 pace with depth. I'm hitting too hard. Oh no, It's a short shot. I feel foolish. Why don't I concentrate? I hope I don't double fault. This is not my day. .. a saw 3 egg :3 233.3: a mass HFHSHE OZOFl-n ZHUJO 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 l 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 APPENDIX J DISCRIMINANT FUNCTION ANALYSIS ON 50 STQ-3 STATEMENTS 187 Table J. Results of Discriminant Function Analysis on 50 STQ—3 Items Itema Wilks' Lambda F Significance L1 .94635 2.84 .0985 L2 .61029 31.93 .0000 L3 .70506 20.92 .0000 L4 .99842 .79 .7794 L5 .43090 66.04 .0000 L6 .98349 .84 .3640 L7 .69082 22.38 .0000 L8 .88889 6.25 .0157 L9 .90844 5.04 .0292 L10 .55717 39.74 .0000 L11 .99959 .20 .8874 L12 .80131 12.40 .0009 L13 .74536 17.08 .0001 Bl .99768 .12 .7344 B2 .76520 15.34 .0003 B3 .90195 5.44 .0238 B4 .96246 1.95 .1687 B5 .97648 1.20 .2777 B6 .98623 .70 .4074 B7 .68057 23.47 .0000 B8 .54122 42.38 .0000 B9 .77480 14.53 .0004 B10 .79083 13.22 .0007 B11 .78603 13.61 .0006 B12 .89842 5.65 .0213 B13 .67680 23.88 .0000 314 .90788 5.07 .0287 BIS .82177 10.84 .0018 B16 .96612 1.75 .1915 B17 .78856 13.41 .0006 188 (table continues) Table J. (Continued) Itema Wilks' Lambda F Significance B18 .66687 24.98 .0000 P1 .99831 .85 .7721 P2 .88862 6.27 .0156 P3 .79475 12.91 .0007 P4 .53333 43.75 .0000 P5 .79699 12.74 .0008 P6 .92356 4.14 .0472 P7 .82179 10.84 .0018 P8 .98765 .63 .4329 P9 .96149 2.00 .1632 D1 .96820 1.64 .2059 D2 .99990 .51 .9436 D3 .99900 .50 .8240 D4 .77179 14.78 .0003 D5 .46222 58.17 .0000 D6 .79048 13.25 .0006 D7 .99452 .28 .6021 D8 .91032 4.93 .0310 D10 .86581 7.75 .0076 D11 .87669 7.03 .0107 8The letter which precedes the STQ-3 item number indicates the game situation category to which the statement belongs: L B Leading in a set Behind in a set Pleased with play Displeased with play REFERENCES 190 REFERENCES Anderson, M. (1981). Assessment of imaginal processes: Approaches and issues. In T. Merluzzi, C. Class, & M. Genest (Eds.), Cogpitive assessment. New York: Guilford Press. Babbie, E. (1979). Theypractice of social research. Belmont, Ca.: Wadsworth. .5 Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review, 34, 191-215. p Bandura, A. (1978). The self-system in reciprocal determinism. American Psychologist, 33, 344-358. Bandura, A., & Adams, N. (1977). Analysis of self-efficacy theory of behavioral change. Cognitive Therapy and Research, 3, 287-310. Beck, A., Rush, A., Shaw, B., & Emery, <3. (1979). Cognitive theory of depression. New York: Guilford Press. Bennet, B., & Stothart, C. (1980). The effects of a relaxation-based cognitive technique on sport performance. In P. Klavora & K. Wipper (Eds.), Psypholggical and sociological factors in sport. Toronto: Publications Division, School of Physical and Health Education, University of Toronto. Bistline, J., Jaremko, M., & Sobleman, S. (1980). The relative contributions of covert reinforcement and cognitive restructuring to test anxiety reduction. Journal of Clinical Psychology, 39, Borkovec, T. (1976). Physiological and cognitive processes in the regulation of anxiety. In G. Schwartz 8 D. Shapiro (Eds.), Consciousness & self-regulation: Advances in research. New York: Plenum Press. Bruch, M. (1978). Type of cognitive modeling, imitation of modeled tactics and modification of test anxiety. Cognitive Therapy and Research, 3, 147-164. L Brunk, J. (1977). Piaget's theory of cognitive and intellectual development. In D. Hamachek (Ed.), Human dynamics in psychology and education (3rd ed.). Boston: Allyn and Bacon. 191 192 '(CaciOppo,.I.,Glass, C., & Merluzzi, T. (1979). Self-statements and self-evaluations: A cognitive-response analysis of heterosocial anxiety. Cognitive Therapy and Research, 3, 249-262. ”/ CaciOppo, J., & Petty, R. (1981). Social psychological procedures for cognitive response assessment: The thought-listing tech- nique. In T. Mesluzzi, C. Glass, & M. Genest (Eds.), Cognitive assessment. New York: Guilford Press. Carmines, E., & Zeller, R. (1979). Reliability and validity assess— ment. In J. Sullivan (Ed.), Series: Qpantitative applications in the social sciences. Beverly Hills: Sage Publications. ‘ Carmody, T. (1978). Rational-emotive, self-instructional and behav- ioral assertion training: Facilitating maintenance. Cognitive Therapy and Research,_3, 241-253. ./ Cooley, W., & Lohnes, P. (1971). Multivariate data analysis. New York: Wiley and Sons. Cooley, E., & Spiegler, M. (1980). Cognitive vs. emotional coping responses as alternatives to test anxiety. Cognitive Therapy and Research, 33 159-166. ..Craighead, L., & Craighead, W. (1980). Implications of persuasive communication research for the modification of self-statements. Cognitive Therapy and Research, 3! 117-134. “’Cratty, B., & Davis, P. (1984). The content of athletes' mental lives: Process decision when considering modifications. In W. Straub & J. Williams (Eds.), Cognitive sport_psychology. Lansing, N.Y.: Sport Science Associates. Derry, P., & Stone, G. (1979). Effects of cognitive-adjunct treat- ments and assertiveness. Cognitive Therapy and Research, 3, 213-221. :, DeWitt, D. (1980). Cognitive and biofeedback training for stress reduction with university athletes. Journal of Sport Paycholpgy, 2, 288-294. Dunkel, L., & Glaros, A. (1978). Comparison of self-instructional and stimulus control treatments for obesity. Cogaitive Therapy and Research, 3, 75-78. Ellis, A. (1979). The rational-emotive approach to counseling. In H. Burks & B. Stefflre (Eds.), Theories of counseling_(3rd ed.). New York: McGraw-Hill. 193 Ellis, A. (1980). Rational emotive therapy and cognitive behavior therapy: Similarities and differences. Cognitive Therapy and Research, 23 325-340. Ericsson, K., & Simon, H. (1981). Sources of evidence on cognition: A historical overview. In T. Merluzzi, C. Glass, & M. Genest (Eds.), Cognitive assessment. New York: Guilford. Ericsson, K., & Simon, H. (1980). Verbal reports as data. Psycho- logical Review, 31, 215-251. ' Feltz, D. (1984). Self-efficacy as a cognitive mediator of athletic performance. In W. Straub 8. J. Williams (Eds.), Cognitive sport psycholagy. Lansing, N.Y.: Sport Science Associates. .p Fuson, K. (1979). The development of self-regulating aspects of speech: A review. In G. Zivin (Ed.), The development of self- regulation through private speech. New York: Wiley. Gallwey, W. (1974). The innerygame of tennis. New York: Random House. / Genest, M., & Turk, D. (1981). Think-aloud approaches to cognitive assessment. In T. Merluzzi, C. Glass, & M. Genest (Eds.), Cpgnitive assessment. New York: Guilford. Girodo, M. (1977). Self-talk: Mechanisms in anxiety and stress management. In C. Spielberger & I. Sarason (Eds.), Stress and anxiety. Washington, D.C.: Hemisphere. Girodo, M., and Roehl, J. (1978). Cognitive preparation and coping self-talk: Anxiety management during the stress of flying. Journal of Consulting and Clinical Psychology, 53, 978-989. , Girodo, M., & Stein, S. (1977). Self-talk and the work of worrying on confronting a stressor. Cognitive Therapy and Research, 3, 305-307. (Girodo, M., & Wood, D. (1979). Talking yourself out of pain: The importance of believing that you can. Cognitive Therapy and Class, C., Merluzzi, T., Biever, J., & Larsen, K. (1982). Cognitive assessment of social anxiety: DevelOpment and validation of a self—statement questionnaire. Cognitive Therapy and Research, a, 37—55 0 ‘fGlogower, F., Fremouw, W., & McCroskey, J. (1978). A component analysis of cognitive restructuring. Cognitive Therapy and Research, 3) 209-223. 194 Gologor, E. (1979). Psychodynamic tennis: You, your ppponent and other obstacles to_perfection. New York: William.Morrow and Company. ’Gormally, J., Sipps, G., Raphael, R., Edwin, D., & Varvil-Weld, D. (1981). The relationship between maladaptive cognitions and social anxiety. Journal of Consultingaand Clinical Psycholagy, 32. 300-301. Gould, D., & Weiss, M. (1981). The effects of model similarity and model talk on self-efficacy and muscular endurance. Journal of Sport Psyaholqu,‘3, 17-29. Gould, D., Weiss, M., & Weinberg, R. (1981). Psychological character- istics of successful and nonsuccessful big ten wrestlers. Journal of Sport Psychology, 3, 69-81. Gravel, R., Lemieux, G., & Ladouceur, R. (1980). Effectiveness of a cognitive behavioral treatment package for cross-country ski racers. Cognitive Therapy and Research, a, 83-89. Hamberger, K., & Lohr, J. (1980). Rational restructuring for anger control: A quasi-experimental case study. Cognitive Therapy and Research, a, 99-102. ‘* Harris, D., & Harris, B. (1984). The athlete'sgnide to sports psychology: Mental skills for physical people. New York: Leisure Press. Hays, W. (1973). Statistics for the social sciences (2nd ed.). Chicago: Holt, Rinehart and Winston. Highlen, P., & Bennett, B. (1980). Predicting performance of elite athletes: The importance of cognitive mediators. In P. Klavora & K. Wipper (Eds.), Psycholgglcal and sociolqgical factors in Sport. Toronto: Publications Division, School of Physical and Health Education, University of Toronto. Hollandsworth, J., Glazeski, R., Kirkland, K., Jones, C., & Van Norman, L. (1979). An analysis of the nature and effects of test anxiety: Cognitive behavioral and physiological components. Cognitive Therapy and Research, 3, 165-180. Hollon, S., & Kendall, P. (1980). Cognitive self-statements in depression: Development of an automatic thoughts questionnaire. Cpgnitive Therapy and Research, a, 383-395. Huck, S., Cormier, W., & Bounds, W. (1974). Reading statistics and research. New York: Harper & Row. \/ Kendall, P., & Hollon, S. (Eds.). (1979). ‘Qagnitive-behavioral inter- 195 Hurlburt, R., & Sipprelle, C. (1978). Random sampling of cognitions in alleviating anxiety attacks. Cognitive Therapy and Research, 2, 165-169. [