CHAPTER 2 Emotions in Sport Current Issues and Perspectives YURI L. HAN1N Selected issues and perspectives on pleasant and unpleasant emotions experienced by athletes and how and why these emotions affect athletic performance are reviewed in this chapter. A balanced view of emotion-performance relationships requires an overview of a sequence involving three groups of individual difference variables: defining characteristics of emotional experiences, antecedents of emotional experiences, and consequences of emotions for athletic performance. Kuhl (1994) used such a sequential framework for description of a theory of action and state orientations, whereas Vallerand and Blanchard (2000) proposed an "antecedents-consequences" sequence for an integrative analysis and review of emotion theory and research in sport and exercise. The chapter is based on an individual-oriented and sport-specific framework grounded in extensive research, the individual zones of optimal functioning (IZOF) model (Hanin, 1995, 1997, 2000). A detailed description of the IZOF model is beyond the scope of this chapter; readers are referred to reviews updating the recent developments of the model (Cerin, Szabo, Hunt, & Williams, 2000; Crocker, Kowalski, Graham, & Kowalski, 2002; Hanin, 2000, 2003, 2004; Raglin & Hanin, 2000; Robazza, 2006; Ruiz, 2004: Woodman & Hardy, 2001). The main emphasis here is on defining characteristics of emotional experiences, their antecedents (determinants), and consequences (outcomes, irnpact). Finally, directions for future research as well as Poetical implications are suggested. 1 would like to express my sincere thanks and deep appreciation '"r 'he thoughtful comments and suggestions from Bob Vallcrand, Knut A. Hagtvet, and Claudio Robazza, who read an earlier draft of this chapter. TERMINOLOGY Terminology issues in emotion research involve attempts to find a more precise definition of emotion (and related affective phenomena) and to provide a detailed description of defining characteristics of emotional experiences. Both aspects are briefly reviewed in the sections that follow. Defining Emotion The definition of emotion remains ambiguous (Vallerand & Blanchard, 2000). It has even become a common practice to state that it is intuitively clear what emotion is, but difficult or even impossible to define. According to Parkinson (1994), there are several ways of approaching the definition of emotion: (a) by giving examples of items belonging to the category of emotion; (b) by looking at the different aspects and components of emotional experience (Crocker et al., 2002; Vallerand & Blanchard, 2000); and (c) by considering how various aspects combine with one another and how they interact to make an emotion episode what it is, and (d) by relating and contrasting it with other psychological functions. It is also possible to examine the dozens of already suggested definitions of emotion and select the one that best encompasses all or most of the research. However, the problem with such an ideal definition of emotion is that it requires a statement of the necessary and sufficient conditions for application of the term, and that is usually not an easy task (Plutchik, 1980). Therefore, an attempt to define emotion is obviously misplaced and doomed to failure. ... To ask today what is emotion is old-fashioned and likely to lead to semantic hairsplitting; to construct systems that unequivocally explain, predict, and make understandable parts of the range of human experience and 31 32 Motivation, Emotion, and Psychophysiology behavior may, in the long run. be the best or only reply. (Man-dler, 1975, pp. 10-11) Interestingly, in current practice, researchers recognize the fact that there is no perfect term and simply sidestep the search for the definition, instead discussing dimensions, categories, and components of emotion (Vallerand & Blanchard, 2000). Additionally, terms describing different affective phenomena (emotion, mood, affect, temperament) are often contrasted (Crocker et al., 2002), although this does not seem to be an effective strategy. Whatever the general definition of emotion proposed, it is important to distinguish among its defining characteristics, antecedents, and consequences (outcomes). Also important is that "we might start not with the aim of explaining emotions but rather with describing a system that has as its product some of the observations that have been called 'emotion' in common language" (Mandler, 1975, p. 4). This is especially true in sport, as is evident in Martens's (1987, p. 51) comment: Sport psychology is theory poor. . . . We have been so eager to test theories of the larger field of psychology in order to confirm our scientific respectability that we have not adequately observed, described, and theorized about our own thing— SPORT. We clearly need to spend more time observing behavior in sport and building our own theories unique to sport. Unfortunately, this concern remains current in sport psychology research, and a need for an accurate and detailed description of emotional experiences is often underestimated or simply ignored. This results in a premature theoretical speculation in the absence of an adequate database (Hanin, 1997; Raglin & Hanin, 2000). To summarize, one option is to continue a search for a more precise definition of emotion; the other option is to focus on an accurate and detailed description of defining characteristics of emotion and relating it to some specific category. Emotion as a Category of Experience Traditionally, emotion as a category is defined as an organized psychophysiological reaction to ongoing person-environment (P-E) relationships. For instance, Deci's (1980 p 85) working definition conveys at least the meaning of emotion: An emotion is a reaction l0 a stimulus event (either actual or - .ned, lnvolves change in (hc v)scera Of the person, ,s experienced subjectively in characterise c ways. expressed through such mea„s as facia, chan u =rcncics-and —* - •zation views emotions as Another working charaCtC^entSi or objects, with their valenced reactions to events, ag < ^ ^ {he e,icjtin„ particular nature being determt^^ & Col]ins, 1988). In situation is construed <^n* dasses of emotions that this approach, there are three d qf the result from focusing on one or and their world: events and their w . ^ rnntpvt evil includes the following sequence: Appratsal ^Context evaluation -> Action readiness Physiological change, expression, action (see Oatley & Jenkins, 1992, for review). In most cases, the definition of emotion as a reaction captures only one aspect of the P-E interaction. The person's response is related to, but still separate from, the environment. Moreover, a descriptive definition of emotion is somewhat limited because it does not include the causal cognitive, motivational, and rational variables and processes involved in arousing and sustaining an emotion (Lazarus, 2000, p. 230). The cognitive-relational motivational theory of emotion elaborates the notion of P-E interaction as applied to stress-related emotions and later to pleasant and unpleasant emotions. To study something as an indivisible unity, according to Vygotsky (1926/1984), it is necessary to find a construct that appropriately captures the characteristics of both interacting elements. In psychology, experience is a relevant construct to study P-E interactions because it reflects a person's attitude toward different aspects of the environment and the meaning of the environment for the person. Experience has a biosocial orientation as every experience !s berre801"011? CXPerienCe °f -das such, nalvs sTSTftaS 3 Unlt °f -Piousness. Thus, the «*■■ * ^~;rr:rse but °n - Emotional exnerien tag. or anticipatecl P-E i2l ? °f paS'' °"t"" taentified at ,eas, tk^T^-^^ ("26/ dominance of an ■ interactions: the pre balance, and the preST^ °^ environment- the P ' organism. These n t" °minance of tne environment over a" tions in sport (Ua ° WCre aPPlied to performance em<> P-E l^7), and it was proposed tl,- between task dem t represented by the relationships 2003, 2004). Fro w ^ * P6rson's resources (Hani"-sport should des"^ h, PerSpective' emotion research » optimal and dysfunor' T^^' and explain an athk"U' ysiunct.oual experiences accompanying ! Emotions in Sport 33 vUlually successful and poor performances. A working def nl|lu,M of experience includes the totality of past and pres-,,„ characteristics that determines the particular quality of a person's performance (Hanin, 2003). ,n the sport context, there are three interrelated types of perft»rmance-related experiences: state-like experiences, or emotional states, as a component of situational, multimodal, and dynamic manifestations of total human func-tioning; traitlike experiences, or relatively stable emotion patterns (emotionality, dispositions, qualities) reflecting a repeated nature of athletic activity; and meta-experiences (awareness, attitudes, preferences/rejections of one's experiences: Mayer & Stevens. 1994), which are lessons learned or reflected experiences in successful and less than successful performances (Hanin, 2004). In contrast to situational states and repeated patterns of experience, meta-experiences reflect how an athlete feels about his or her past, present, or anticipated emotional experiences and the perceived effects of these emotional experiences on performance or general well-being. For instance, an athlete may feel nervous and uncertain prior to a competition. That characterizes his or her situational emotional state as triggered by a specific meaning of the particular situation for this athlete. On the other hand, feeling nervous can be a typical (repeated) pattern of this athlete's emotional response in similar situations. Therefore, in this particular case, trait competitive anxiety would indicate how often the athlete experiences elevated anxiety and feels nervous, tense, or apprehensive prior to or during competition. However, an athlete's meta-experience (attitude to experiencing a high level of competition anxiety and awareness of its helpful or harmful effects on performance) is even more important to estimate. Meta-experiences are formed when athletes (and coaches) spontaneously and deliberately reflect on the conditions leading to their successful, and less than successful, performances. Meta-experiences determine an athlete's perception and a choice of coping and self-regulation strategies, and therefore should be a major target of interventions. Interestingly, most research in sport psychology during ihe past 2 decades has focused mainly on situational emotional states (such as competition anxiety) and relatively stable emotion patterns (e.g., trait anxiety). Meta-experiences in sport, although undefined as a separate parameter (Hanin, 2003), were actually implied in the assessment of optimal and dysfunctional zones of emotion intensity (Hanin, 1978, 1986; Hanin & Syrja. 1995) and in the ratings of "directional" anxiety (or perceived impact) on performance (Jones, 1995). On the other hand, in practice, emotion regulation is °ften based on reframing an athlete's attitude toward specif- ic emotional experiences. For instance, it is difficult to imagine how an athlete can constructively use high anxiety without a positive attitude and expectation of its helpful effects. In other words, meta-experience adds a special meaning and a new quality to perceived situational state, which is interpreted (or reinterpreted) as facilitating or debilitating. Therefore, the role of meta-experiences as determinants of appraisal and coping processes should be reemphasized, especially in intervention studies. Based on Vygotsky's suggestion, emotion is construed not as a reaction, but as experience (situational and repeated) and meta-experience reflecting the dynamics of P-E interactions. DEFINING CHARACTERISTICS OF EMOTION EXPERIENCE A comprehensive analysis and understanding of emotion experiences in sport requires an accurate description of their basic dimensions or defining characteristics. What are these basic (i.e., sufficient and necessary) dimensions? Apparently, emotion experiences are complex phenomena requiring multidimensional characterization. For decades in emotion research, typical dimensions were valence (i.e., hedonic tone) and intensity. Both were used in conceptualizing global emotion content (pleasure/displeasure and high and low activation). On the other hand, historically, emotion components have been characterized by three parameters derived from measurement methods rather than from the conceptualization of emotion dimensions. These include physiological concomitants, introspective (verbal) self-reports, and behavioral observation (Eysenck, 1975). From this perspective, typical dimensions are emotion intensity, emotion valence, and emotion manifestation as assessed by cognitive labels, bodily response, and behavioral displays (expression or suppression). A need to go beyond these widely accepted dimensions to capture a more complete picture of emotional experiences is clearly indicated (Hanin, 1995, 1997, 2000, 2003). In the sections that follow, a brief description of the five basic dimensions characterizing emotion experiences is provided. Multidimensionality of Emotion Experiences An alternative multidimensional approach was proposed in the IZOF model (see Hanin, 1997, 2000, for a review). It was derived from the method of bases developed for the systems description of complex phenomena (Ganzen, 1984). In the systems description, a multitude of elements of the object under investigation is contrasted with the elements of the basis (the logical foundation). Ganzen, having analyzed the descriptions of different objects and phenomena, proposed 34 Motivation, Emotion, and Psyehophysiology that "spatiality, time, information and energy were the basic characteristics of any object that typically functions as their integrator" (p. 44). These separate concepts (space, time, energy, information, and a substrate) were suggested as a conceptual basis (pentabasis, or a five-element foundation) to integrate existing concepts and empirical research findings. This descriptive framework makes it possible to (a) examine the completeness of description of the phenomenon, (b) better organize the components, (c) compare different descriptions, and (d) discover the similarity in the objects or phenomena of different natures (pp. 41-42). This approach has been theoretically substantiated and empirically validated in the systems descriptions of psychological subdisciplines, general characteristics of the nervous system, and the description of human personality and individuality (Ganzen, 1984). In the sports setting, the pentabasis and the idea of systems description were used in the longitudinal study of communication patterns in top sport teams (Hanin, 1980, 1992), in sports career and athlete crisis research (Stambulova, 2000), and in investigations of performance-related emotions (Hanin, 1993, 1995, 1997,2000). In its current form, the IZOF model posits five basic dimensions that capture defining characteristics of emotion experience as a component of different psychobioso-cial states related to performance (Hanin, 2000, 2003). I argue that emotional experience is always manifested in some form (subjectively perceived or observable); it has specific content (or quality); it is characterized quantitatively by its intensity and as a process that unfolds over time (Folkman & Lazarus, 1985) in a particular context. Thus, the multilevel and system description of emotion as a component of performance-related states should include at least five interrelated dimensions: form, content, intensity, time, and context. Three of these dimensions (form, content, and intensity) describe the structure and function of the subjective emotional experiences and meta-experi-ences; time and context characterize dynamics of performers' subjective experiences in a specific social setting. Actually, these five basic dimensions include traditional emotion components (implied form, valence, and intensity) and provide a tool for a systems description of emotional experiences (for more detail, see Hanin, 1997, 2000, 2003, 2004; Robazza, 2006). The following sections focus mainly on emotion form, content, and intensity. Situational Emotion and Nonemotion Experiences An athlete's performance state manifests itself in the form mliv"' WhiCh.COnsists of "even basic components or bodi behaC°gnitre' em°tl0nal (affective>< motivational, bodily, behavora., operational (action tendencies), and u ,n ,997 2000, for a review). Fron, communicative (see emotional experience (e.g.. this perspective, situation^ ^ ^ ^ psychobiosociaj anxiety or anger) state related to nQfm^ZZc^rch focuses on emo-Current individual-one components Gf perfor- tional, motivational, and effects. Recem mance state and tne describe perfor- emptrical evidence ^ use their own vocab. mance-relatedexpertence^ athletes ulary of idiosyncratic labels, nu ' , v „«rfnrmance-related states usualh vocabulary describing performance i ..." includes not only self-generated emotion words but also labels describing nonemotion experiences: cognitive, motivational, bodily, motor-behavioral, operational, and communicative (Hanin, 1997; Hanin & Stambulova. 2002; Ruiz & Hanin, 2004a, 2004b). For instance, Hanin and Stambulova examined emotional experiences prior to. during, and after personally best and worst competitions in 85 skilled Russian athletes using a metaphor-generation method. Each athlete had to complete a sentence, "Prior to my best competition I felt like . . . ," that generated a metaphor (e.g., "I felt like a tiger") as a symbolic representation of a feeling state. Completing a paraphrased sentence, "In other words, I felt myself. . . ," elicited an interpretation (e.g., "I felt strong and focused") of an athlete's state as symbolized in the metaphor. Then athletes generated metaphors and interpretative descriptors for competition situations during and after performance. The same procedure was repeated to describe how they felt prior to, during, and after worst-ever competitions. These six situations elicited 510 idiosyncratic and functional!) meaningful metaphors and 922 interpretative descriptors. ttinn ' metaph°rS and descriptors reflected high e " competition and low action readiness in worst-f»vp.r ~~ ■ . different m„t V COmPetition. Athletes also used h1« pT t;°rdescdbe'symboiic^th- -eaning'of tnSe si t ^ perfo™ance "s accomPLying ^Uatl°nS Chan^. Interestingly, th, emotional expene„ynkatlc ^ descdbed "0t onb of nonemotion com™ ^ mUltiple connotation* Similar finding °f the P*ychobiosocial ^ accompanyinoll! °n sel ^generated metaphors and studying a sam Jretatlve descriptors were obtained in Hanin, 2004b) ? l°P Spanish karate athletes (Ruiz ^ fined, ius noTtl0n' ^ * COncept' remains largely unctions and no SUrpnSln§ that distinctions between em'1" especially in are betimes not quite ck-a- assessments. For instance, an inspection ^ the 10 global affect scales described by Watson and Te „(1985) shows thai some of the items are 'concern,, n foolty and WOUld "°l bc COnsid^ -""Hons hy ni " centered theorists" (Lazarus, 2000, p. ^ '1 w0|,!s. emotion descriptors in existing emotion scalesI , represen1 not only "pure" emotions, bu, also nonemot on components ol a state (cogn.tive, motivational bodilv and behavioral). Apparently, research-wis^ i, js important clearly distinguish among emotion, nonemotion and b, ," derline modalit.es ol a state (La/arus, 2000) From the applied perspective, however, a more holistic description of the performance-related state, including emotions and nonaction experiences, could be equally important and sometimes perhaps even more appropriate. Recently, Robazza, Bortoli, and Hanin (2004) showed thai athletes are well aware of several nonemotion modalities of their performance state (motivational, bodily, sensory-motor, and behavioral), [n another study (Hanin, 1999), seven positively toned items (motivated, willing, desirous, hopeful, keen, daring, and interested) and seven negatively toned descriptors (unmotivated, unwilling, reluctant, hopeless, bored, compelled, and uninterested) discriminated quite well the motivational states of 29 highly skilled ice hockey players before their successful and less than successful games (see Figure 2.1). In contrast, motivational domains in this sample had multiple and diverse connotations. Table 2.1 provides a summary of responses of these players to a question about what motivates (and what does not motivate) them before the game. rablc2.I Enhancing Ice Hotkey Players Emotions in Sport 35 and Detrimental Motivational Domains for Enhancing Motivational "I'm motivated if. .. " Focus on: • Winning • Fighting • Doing my best ■ Learning Ice hockey: ' My serious hobhy • My future profession • My life Domains Feeling state: • Self-confident • Trust myself • Enjoying the game • Psyched up Opponent: • Tough • Good • Strong Detrimental Motivational Domains "I'm not motivated if. . . " Preparation: • Insufficient recovery • Poor shape • Poor planning Outside sport: • Family • School • Other concerns Feeling slate: • Too tired • Health problems • Dissatisfied • Too satisfied Opponent: • Too easy • Clearly weaker Our game: • Important • Challenging • Tough • Well started Own team: • I play for my team • f work for team's success • Good climate in the team Our game: • Too easy • "Meaningless" • Nothing works • Clearly lost • Bad start Own team: • Repeated losses • Poor team climate Note: N =29 Finnish ice hockey players. Adapted from Emotions in Hockey, by Y. L. Hanin, May 2000, paper presented at the HHF International Coaching Symposium: Building a Hockey Base for the 21 st Century, St. Petersburg. Russia. Adapted with permission. 10 7 4 1 -2 1 4— \ S*^ \xl "■ ■ ■ ■ \ —--' I X ^ 1-4-Good Game -^B^Game ] 2.1 Individualized motivational prohle o h J Sayers W-29). Adapted from "Sports-Spec, .c Em Rational Profiling: An Individualized Asscssn rd «J 238-240), by Y. Hanin, in Psyrlu^S | ^ piltk (Eds.), ;y«<^v of Life, V. Hosek. P. Tiling and psy Proceedings or the l()th European C on,. j(y J**: Part I, Prague, Czech Republic: Charles 18s' Adapted with permission. Numerous athlete-generated bodily descriptors are examples of another component in the form dimension. These idiosyncratic bodily labels included different experiences located in face, legs/feet, arms/hands, neck/shoulders, and stomach (see Table 2.2). Also mentioned were characteristics of movements, heart rate, and feeling thirsty, hungry, cold, and pain (Robazza, Bortoli, et al., 2004). Interestingly, these symptoms are more diverse compared to researcher-generated items, for instance, in the Competitive State Anxiety Inventory (CSAI-2). Future research might identify idiosyncratic bodily descriptors of different emotional experiences related to successful and poor performances across different sports and groups of athletes. Although "reading the players" is an important social psychological skill for a coach, especially in team sports, behavioral indicators of specific emotional experiences have not yet become a focus of systematic studies in sport psychology. Several attempts to examine this modality suggest that coaches and athletes are well aware of the behavioral symptoms of certain emotions. For instance, in an unpublished 36 Motivation, Emotion, and Psychophysiology Face • Tense/relaxed • Nervous lies • Yawns • Dry mouth Movements • Energetic • Vigorous • Sharp • Smooth • Slow • Stiff Legs/feet • Tense • Loose • Cold Heart rate • Perceived • Irregular • Accelerated Arms/hands • Tense/relaxed • Sweaty/cold Feeling • Fresh • Thirsty • Hungry/no appetite • Exhausted/tired • Cold/warm • Sweating • Urinary pressure • Lightness Neck/shoulders Tense Pain • Physical pain • Headache • Back pain • Stomachache • Lack of pain Adapted from Emotions in Hockey by Y. L. Hanin, May 2000. paper presented at the 1IHF International Coaching Symposium: Building a Hockey Base for the 21st Century, St. Petersburg, Russia; and "Pre-Competition Emotions, Bodily Symptoms, and Task-Specific Qualities as Predictors of Performance in High-Level Karate Athletes," by L. C. Robazza, Y. Bortoli, and Hanin, 2004, Journal of Applied Sport Psychology, 16(2), pp. 151-165. Adapted with permission. exploratory study, Hanin (2005) asked 16 ice hockey coaches to describe behavioral markers of a player who feels self-confident. According to these coaches, such a player looks purposeful, relaxed, calm, certain, focused, determined, happy and willing to go on ice. His body language is active. He stands up tall; his nose is not facing the ground; his voice is very sure; he radiates energy; he smiles and talks but stays focused; he looks forward to the situation; he enjoys playing (expressive movements); he makes eye contact with the coach and does not rush. In contrast, a player with low self-confidence prior to the game, is silent; thinks a lot; wants to be alone; sometimes talks too much to forget the game; tries to relax by laughing; worries a lot; asks when to go on the ice. [italics added for emphasis] These coaches were also able to describe, in much detail, observable behaviors of the players who feel high or low anxiety, complacency (satisfaction), and anger. There is clearly a need for development of behaviorally anchored scales enabling controlled observation of athletes' displays (expression or suppression) of emotional experiences prior to, during, and after successful and poor performances. In team sports, the major focus and concern of the coach are the emotional states of the goal keeper and the key players (leaders and subleaders), who affect the emotional dynamics of the entire team. Emotion Content Emotion content as a qualitative characteristic includes such general categories of emotional experiences as positive- negative (Russell, Weiss, & Mendelsohn, 1989; Watson & Tellegen, 1985), functionally optimal-dysfunctional (Hanin, 1978, 1993), and facilitating-debilitating (Alpert & Habcr. 1960; Jones, 1991, 1995). Therefore, content is one of the basic dimensions in the systematic study of emotional experiences. It is difficult to imagine an emotion without a distinctive content and intensity (Lazarus, 2000). Both quality and intensity determine the functional impact of emotions on performance and well-being. Two traditional approaches to categorizing emotion content are the dimensional (global affect) approach and the discrete (basic) emotion approach. The global approach emphasizes pleasantness-unpleasantness (valence or hedo-nic tone), tension-relaxation, and quiescence-activation (Russell, 1980; Watson & Tellegen, 1985). The discrete emotion approach centers on discrete categories of emotion based on their qualitative content (anxiety, anger, joy. etc.) and claims that there ^ dusters Qf J^J d.s. crete emotion syndromes (Lazarus, 2000). notr07ir;;UThion researchers embrace thc ^ea that there lasetof "h " »Sti" * together with th . basic emotions such that they. <«*. o^ST^ acc°unt for a11 emoti°ns section is that any i t'0fh 1 ^ P" 25)" in§ from 3 fSnin '1Stofbasic (discrete) emotions, rang- (Izard), and e^tion .abe.s propo^tv K ( "} ^ e'ght differed approaches t« mVeStiSators representing f°und that, au inan ,u ° emot'on research. It wa* all, there were 47 .abels of basic emotion* Emotions in Sport 37 (with ;- negativel) toned and 15 positively toned emotion descriptors). The most selected emotion labels were fear y\i) researchers), anger (18). sadness (9). and disgust (7): 23 labels were proposed only once, and 10 labels were selected twice (.see Table 2.3). Although any list of discrete emotions is arguable, at least two important aspects were clearly identified by Lazarus (2000). First, the list should include both negativ e-h toned emotions (.e.g.. anger, anxiety, fright, sadness, guilt, shame, envy, jealousy, disgust) and positively toned emotions (relief, hope, happiness/joy. pride, love, gratitude, compassion). Second, regardless of the exact list, "a primary empirical and theoretical concern is to identify the most important emotions, their distinctive characteristics, antecedent causal variables and consequences, and how they might influence competitive performance in sports"" (Lazarus. 2000. p. 232. italics added). In competitive and high-achievement sports, the most important emotions are usually personally relevant, task-specific, and functionally helpful or harmful emotions really experienced by athletes. This assumption has received strong empirical support (Hanin. 1997. 2000. 2004; Robaz-za. 2006) and is based on the notion that "'under similar environmental conditions, people perceive themselves differently, think differently, cope differently, and experience and display emotions differently" (Lazarus, 1998, p. 213). Thus, the functional importance of emotional experiences is associated with their goal relevance and with the extent Table 2.3 Basic Emotions: Frequencies of Label Selection Fear(19) Anxiety (2) Pain (1) Anger (18) Curiosity (2) Panic (1) Sadness(9) Elation (2) Pity (1) Disgust (7) Enjoyment (2) Pride (1) Joy (6) Expectancy (2) Resignation (1) Happiness (5) Loneliness (2) Sleepiness(1) Interest (5) Rage (2) Sensuous comfort (1) Surprise (5) Contempt (2) Sex-lust (1) Lave (4) Appetite (1) Shock (1) Pleasure (3) Grief (1) Subjection (1) Satisfaction (3) Acceptance (1) Succor(1) Shyness(3) Amazement (1) Tenderness(1) Distress (3) Anticipation (1) Tension (1) Shame (3) Boredom (1) Want (1) Guilt (2) Despair (1) Wonder (1) Sorrow (2) Quiet (1) Note: N = 23 researchers. Positively toned emotions are in italics. Adapted from Emotions in Hockey by Y. L. Hanin, May 2000, paper presented at the IIHF International Coaching Symposium: Building a Hockey Base for the 21st Century, St. Petersburg, Russia. Adapted with permission. that each athlete is able to perform up to his or her potential using effectively available resources. In contrast, the usual laboratory study of emotion assumes that if the stimulus conditions are equal for all subjects, then the average of all subjects' responses best represents the group for the variable measured. Implicit in this assumption is the idea of equivalent life and performance histories, which obviously cannot be met in studies with humans. Lacey (1967) has demonstrated that different subjects tend to respond by activating different major physiological response systems, and that within any large group of subjects, several types of responders always exist. Obviously, this is true not only for bodily responses, but also for emotional experiences described by athletes' self-generated idiosyncratic labels (see Hanin, 2000, for a review). Idiosyncratic Emotion Content To identify person-relevant and functionally important emotional experiences, the IZOF model proposes that athletes use their own vocabulary of self-generated idiosyncratic labels. These self-generated emotion labels describe athletes' subjective pleasant and unpleasant experiences prior to (or during) their successful and poor performances. The implication is that success-related experiences are helpful for (or at least do not disturb) an athlete's performance, whereas failure-related experiences are detrimental (harmful) for individual performance. Although the main emphasis of the IZOF model is on emotion effects on athletic performance, the functionality-dysfunctionality of emotions is not limited to perceived (anticipated) helpful/harmful effects on performance. For instance, the functionality of emotions can be based on anticipated emotion effects on postperformance recovery (Hanin, 2002), performance-induced injuries (Devonport. Lane, & Hanin. 2005; Wurth & Hanin, 2005). or an athlete's general well-being (Diener, 2000). Moreover, empirical findings suggest that the functionality of emotions relevant with respect to one criterion, for instance, performance, is not necessarily relevant for other outcomes, such as leisure quality, postin-jury recovery, or general well-being in healing or educational settings. In other words, in each particular setting, functionality-dysfunctionality should be clearly specified as a set of intrapersonal, interpersonal, health, or well-being consequences (see Oatley & Jenkins, 1992. for a general discussion of emotion function and dysfunction). In the IZOF approach developed for the high-achievement setting, emotion content is conceptualized within the framework of two interrelated factors: hedonic tone, or valence (pleasure-displeasure), and performance functionality 38 Motivation, Emotion, and Psyehophysiology (optimal-dysfunctional effects on performance processes and outcomes). Both factors reflect qualitatively different aspects of emotional experiences related to individually successful and poor performances (Hanin, 1997). Selected idiosyncratic emotion labels are classified into one of the four global emotion categories derived from hedonic tone and performance functionality: pleasant and functionally optimal emotions (P+), unpleasant and functionally optimal emotions (N+), pleasant and dysfunctional emotions (P-), and unpleasant and dysfunctional (N-) emotions. Optimal (P+ and N+) emotional experiences accompany successful performances, whereas dysfunctional (N- and P-) emotional experiences are usually related to poor performance. These four emotion categories provide an initial structure that is sufficiently broad and robust to generate a pool of idiosyncratic, individually relevant, and task-specific emotions experienced by athletes prior to, during, and after their successful and less than successful performances. It is important that athlete-generated labels describe idiosyncratic and experientially grounded emotions. Moreover, the individualized framework provides an opportunity for athletes to reflect on and report their most significant pleasant and unpleasant emotional experiences related to their individually successful and poor performances. Self-generation of idiosyncratic personally relevant labels, assisted by an emotion stimulus list (Hanin, 1997, 2000, 2003; Robazza & Bortoli, 2003), is a feature that makes the IZOF approach different from both global affect and discrete emotion approaches. In the individualized approach, the pleasure-displeasure distinction is similar to a global dimensional approach, which, however, does not have the functionality-dysfunc-tionality distinction. Additionally, the four-category global framework does not limit, in any way, selection of the most appropriate idiosyncratic emotion descriptors. Therefore, athletes reconstruct their performance-related experiences by generating their own idiosyncratic labels. They are not forced to squeeze their unique subjective experiences into researcher-generated descriptors of preselected discrete emotions (anxiety, anger, joy, etc.). Moreover, self-generated labels reflecting an athlete's perspective, when aggregated across athletes and sport events, identify prototype (most often selected) emotional experiences that can be recategorized using a selected discrete emotion framework (Hanin, 2000, 2004; Hanin & Syrja, 1995" Robazza, 2006; Ruiz & Hanin, 2004a) It is reasonable to ask about the extent to which the content of athlete-generated emotion labels are similar to (or different from) researcher-generated emotlon ]abe,s ^ , e Conversely, how are sell-in existing standardized scale- to the exist-generated idiosyncratic emotio ^ questions. ing lists of discrete etnouonshou]d be con. emotion experiences of inu oriented emotion scales trasted with standardized hoW athietes feel before, that are currently used to desc popular scales during, or after performance. ^ Spieiberger, Gorsuch, developed in nonspor,' se'™« i( Anxie,y Inventory Mood State (POMS,, and and Te^ens < > :i:r:deAr.enS, (1990) CSAI-2 and Smith, Smoll, and Schutz s (1990) Sport Anxiety Scale (SAS). One problem with most group-oriented scales is that they use a pool of researcher-generated items with "fixed-emotion content (global or discrete). These similar emotion items usually imply the same psychological meaning of emotion descriptors for all athletes. However, in most cases, it is not known to what extent emotion content assessed with the group-oriented scales reflects emotion content really experienced by individual players in their successful and poor performances. Two studies involving 50 skilled soccer players and 46 ice hockey players compared the content of emotion items in STAI, POMS, PANAS, and CSAI-2 scales and individual emotional experiences assessed by athlete-generated labels (Syrja & Hanin, 1997, 1998). The findings revealed that 80% to 85% of self-generated emotion labels were not included in the selected standardized scales. In other words, the scales with researcher-generated items did not assess 80% to 85% of the emotional content of athletes' performance-related subjective experiences. These findings received additional empinca support in another study involving Spanish elite karate athletes who expressed individual preferences in the el ction of idiosyncratic labels describing their anger state, of varying intensity (RU1Z & Hanin, 20§04a) In another studv (Rui? & u • ~ emotion labels general b! ^Tlf^' athletes wer. "erated by 16 high-level Spanish karate athletes were compared with the list of 15 discrete emotions proposed by Lazarus f2nnm t J ais"ete emo tion profiling, these athW individualized emo- V^^ftnSoS^ §enerated % idio^ic' interpretative labetdescrih u "g mCtaph°rS *** ^ ing, and after ^ prior to, dur- expected, self-generated inL ^ performances- AS ^werehigMy^d^;ntriIsedtatlve emotion derp^ self-generated idiosynemic 1 k ? C°ntext-sPecific- Thesi' •osyncratic labels were related to three Emotions In Sport 85 plMSant discrete emotions (happiness, pride, and relief) alK1 three stress-related unpleasant emotions (anger, anxi L-1\. and sadness). Additionally, athletes" experiences in worst performance were related to fright and shame. Inter ostmgly. the athletes" self-generated labels had no content overlap with seven other discrete emotions (love, hope, compassion, gratitude, envy, jealousy, and guilt) proposed in La/.arus (1991, 2000). These findings suggest a speci ficity of emotion content in high-achievement settings, especially it the emphasis is on such extreme and qualita liveh different situations as success and failure. Pure or Mixed Emotions Systematic assessment of the idiosyncratic emotion content of athletes' experiences provides an answer to the question about pure and mixed emotions. Most of the research in sport psychology during the past decades has focused on selected stress-related emotions, such as anxiety. As a result, the complex picture of actual emotional experience was oversimplified and incomplete at best. Research into pleasant and unpleasant idiosyncratic emotions has made it increasingly clear that in real-life situations, athletes' experiences are better described by mixed rather than pure selected emotions (Oiener & Emmons, 1985; Gould & Tuffey, 1996; Hanin. 1997. 2000. 2003; Hanin & Syrja. 1995; Jones & Hanton. 2001; Morgan. 1984: Plutchik. 1980; Schimmack, 2001). To illustrate this notion, idiosyncratic emotion labels generated by a junior international-level tennis player describing his emotional experiences prior to, during, and after his best and worst games are presented in Figure 2.2a and 2.2b. Prior to his best-ever game (Figure 2.2a). the player felt high intensity of pleasant optimal emotions (P+): He felt highly determined, confident, excited, dynamic, and comfortable. He also felt moderately aggressive, alarmed, and somewhat uncertain (N+) at the same time. Moreover, his unpleasant dysfunctional emotions (N-; nervous, afraid, worried, and intense) were of low intensity. This pattern was similar during that game, except that he felt alert and quick but not too excited and had no premature satisfaction. In contrast, prior to his worst-ever game (Figure 2.2b). this player fell highly nervous and worried (N—), and these experiences were even more intense during the game. Interestingly, at the same time, his optimal pleasant emotions prior to and during the game were of moderate and low intensity, respectively. II only the anxi ety level in this player in his best and worst games were measured, the entire profile of his emotional experiences and their impact on his performance would be missed. Clusters of emotion content and intensity change from pregame to midgainc and postgame situations for this player. Because his emotional experiences are related to different aspects of the environment, they are, again, better described by a cluster of mixed emotions rather than by a few pure or discrete emotions. Mixed emotions reflect a set of different domains that are perceived by an athlete in a particular perlonnance situation or significant events outside sport. Interestingly, a similar mixture of motivational domains was established in ice hockey players describing what can motivate or de-motivate them before the game (see Table 2.1). Future reseauh in sport psychology should focus on mixed pleasant and unpleasant emotions representing actually experienced slates rather than pure emotions. Also, the effect of discrete emotions, such as anxiety or anger, should be analyzed in the context of other, potentially related emotions. Finally, although mixed emotions certainly represent one important aspect of performance-related experiences, another aspect emerged in the analysis of labels generated by athletes. It was revealed that there are emotion mixtures and mixtures of nonemotion components (alert, energized, motivated, determined) of the psychobiosocial state (Hanin, 1993, 1997). Similar supporting data were obtained when standardized normative scales were contrasted with idiosyncratic emotion descriptors generated by athletes (Hanin, 2000; Syrja & Hanin, 1997, 1998). Developing an empirical typology of "emotion mixture" seems like a promising future direction in emotion content research in sport (Diener & Emmons, 1985; Hanin, 1993, 1997; Schimmack, 2001). Emotion Intensity Emotion intensity is one of the most important dimensions; together with emotion content, it determines the effect of emotion on athletic performance. Numerous studies focused on the link between intensity of anxiety and performance outcomes in different athletes. However, assumptions that the optimal level of anxiety intensity in all athletes should be either moderate (U-inverted hypothesis), high (drive theory), or low (quiescence model) did not receive much empirical support. In most cases, the curves describing, for instance, the shape of anxiety-performance relationships in the zero-maximum range of intensity (from sleep to extreme excitement) were tentative at best. Most of these curves were based on two or three cross-sectional comparisons of anxiety levels in groups of athletes (Landers. 1994). These data usually did not include the entire working range of intensity because under laboratory conditions it is quite a challenge to manipulate the intensity level along the entire (a) Emotion dynamics in a "best-ever" game Poor Successful Poor i- ľ g+ F ,. Before • Nervous 2 • Alarmed 3 the game • Afraid 2 • Aggressive 5 • Worried 1 • Intense 1 • Determined 9 • Confident 8 • Excited 8 • Comfortable 9 • Dynamic 7 • Satisfied 2 • Sure 3 • Easy 3 During • Tense 2 • Aggressive 4 the game • Nervous 2 • Unsafe 4 • Intense 1 • Uncertain 2 • Anxious 2 • Alarmed 4 • Confident 9 • Determined 10 • Alert, quick 9 • Comfortable 7 • Exhausted 1 • Concerned 1 • Comfortable 8 • Happy 8 • Restless 1 • Inspired 7 • Satisfied 9 • Determined 8 • Easy 4 • Focused 1 • Pleased 2 After the game (b) Before the game During the game After the game Poor Emotion dynamics in a "worst-ever" game Successful N + P + • Nervous 9 • Afraid 7 • Worried 7 • Intense 6 • Tense • Nervous • Insecure • Anxious • Angry Alarmed 4 Aggressive 6 Aggressive 3 Unsafe 6 Uncertain 5 Alarmed 5 Determined 5 Confident 4 Excited 3 Comfortable 3 Dynamic 2 Confident 2 Determined 1 Alert, quick 3 Comfortable 1 Exhausted Unhappy Irritated Dissatisfied Restless Angry 6 Annoyed 4 Concerned 3 Comfortable 1 Inspired -| Determined 4 Poor P- • Satisfied • Sure • Easy 2 1 2 * Relaxed 3 • Easy 2 Satisfied Easy Pleased Figure 2.2 Emotional dynamics in an international-level tennis player in best (a) and " Related Emotional States in Sport: A Qualitative Analysis" [48 paragraphs, Online i (b) ?ames. Adam a r Qualitative Sozialforschung, 4{\), available from http://www.qualitative-research °Urna'l, by Y. L H-in I m "Performance ■net/^s-texte/l-03/i.o3hänin Uary 2°°3' Forw" permission. 40 ary e-htm. Adapted with Emotions in Sport 41 range of intensity. Interestingly. Yerkes and Dodson (1908) were not sure if the levels of intensity of the stimulus used in their experiments with mice were "most favorable." The problem becomes even more complicated when separate and interactive effects of different components of anxiety, or multiple pleasant and unpleasant emotions, are examined. Applied research and practice in high-achievement sport, however, require a more Individualized approach that can predict an individual performance. One strategy to solve the problem was proposed by Hanin (1978. 1995. 1997. 2000). who argued that it is unproductive to focus only on actual anxiety and corresponding levels of performance, matters that are difficult to compare across athletes. For instance, what one athlete would consider a good or even excellent performance could be perceived by another athlete as poor. Therefore, the emphasis should be on analysis of past performance history and estimation of intensity of emotions accompanying individually successful and unsuccessful performances. Because the "moderate anxiety for all" assumption did not work in practice, a more intraindividual focus and individualized criteria in the evaluation of current anxiety intensity were needed (Hanin, 1978. 1995; Raglin & Hanin. 2000). Several studies have reported the percentage of athletes performing their best when experiencing high, moderate, or low anxiety (see Jokela & Hanin. 1999). The distribution of athletes in these categories was surprisingly well balanced across different studies: high (Af=34.2; 26% to 50%), moderate (M = 34.6: 22% to 44%), and or low (M=35; 25% to 48%). Moreover. Jokela and Hanin (1999) were unable to identify a single study in their metaanalysis that demonstrated that different athletes had the same (or similar) optimal levels of anxiety. The individual-oriented strategy proposed by Hanin (1978, 1986. 1989) to predict the effects of anxiety on athletic performance emphasized a need to analyze an athlete's past performance history to identify emotions accompanying individually best performances. The main emphasis in this approach is on predicting individual performance by contrasting, for instance, current anxiety level with the previously established success-related anxiety level (high, moderate, or low). The concept of zones of optimal functioning (ZOF) initially proposed in precompe-tition anxiety research was a tentative optimal range of intensity scores predicting individually successful performance. Later, the ZOF concept, extended to pleasant and unpleasant emotions, was later termed IZOF (individual zones of optimal functioning) to emphasize the within-individual focus of the model (Hanin. 1995, 1997, 2000). Probability of successful performance was high when current precompetition anxiety was near or within the previously established individually optimal intensity zones. When precompetition anxiety fell outside the zones (i.e., higher or lower), individual performance usually deteriorated. The interest in individually oriented optimal zones of anxiety intensity reflected the fundamental fact that each athlete has a unique set of resources that are situa-tionally available (or unavailable) for coping with the demands of an environment. Recently, similar results were obtained in studies of optimal and dysfunctional effects of situational anger on athletic performance (Ruiz & Hanin, 2004a. 2004b). There were several advantages of the individualized approach to precompetition anxiety based on the realities of high-achievement sport and an accurate description. First, the step-by-step methodology for establishing the IZOFs was proposed. Second, an athlete's past performance history was considered, and individually optimal anxiety level and zones were established. Third, testable predictions of individual (and group) performance based on current anxiety and IZOFs were available. Fourth, the approach was empirically tested using different anxiety measures (STAI, CSAI-2, POMS, and the Body Awareness Scale; Koltyn & Morgan. 1997: Wang & Morgan, 1987) across different samples, different sports, and different countries. Numerous studies provided strong empirical support for the approach and the recall methodology of assessing optimal levels and zones of individually optimal anxiety (Hanin, 1995; Jokela & Hanin, 1999. meta-analysis). However, initially, the IZOF anxiety model focused on precompetition anxiety as a discrete stress-related emotion syndrome with "fixed" emotion content, and the main emphasis in the IZOF anxiety research was on identifying the individually salient intensity of state anxiety (Raglin & Hanin, 2000). The IZOF notion was proposed as an experience-based, individualized criterion to predict individual performance. The concept was derived from observations of real emotional experiences of athletes that were optimal in individually successful performances. When an athlete's anxiety was out of the optimal zone, his or her performance clearly deteriorated. Empirical findings consistently demonstrated high interindividual variability of optimal precompetition anxiety across different samples of elite and competitive athletes (Hanin, 1978. 1995; Raglin. 1992; Raglin & Hanin, 2000; Raglin & Turner, 1993). Therefore, the IZOF concept became a guiding principle in . and Psychophysiology 42 Motivation, Emotion of individual-oriented performance predictions based on the IZOF anxiety hypothesis, many questions arise: Do optimal and dysfunctional intensity levels and zones change during a season? And if they do, how are these changes related to an athlete's available resources and readiness for a competition? Does the accuracy of recall change with an athlete's increased self-awareness? What is the validity and reliability of the empirical method of intensity zone estimation (direct observations)? Can it be used without a recall method? How are the intensity levels and zones related to the optimal and dysfunctional impact of emotion on performance? How and why, for instance, is high anxiety helpful or harmful to individual performance? Finally, how can we enhance the accuracy of intensity zone estimation based either on categorical (either in or out of the optimal zone) or continuous measures along the entire working range of intensity? These and other questions provide directions for future work. For instance, the empirical (direct) method of estimation of intensity zones consists of repeating actual assessments in several successful and unsuccessful competitions, plotting emotion intensity levels, and evaluating the distribution of optimal intensity scores (Hanin, 2000, p. 164). Traditionally, optimal intensity levels and zones are based on either the mean ±0.5 standard deviation range or on the interquartile range (IQR), which includes the range of scores from the 25th percentile to the 75th percentile. The IQR is one of several interpercentile measures of variability that tell how the middle 50% of the distribution is scattered. The clear disadvantages of the direct assessment method are that it requires many data points, it ignores an athlete's past performance history, and it is usually limited to pre- and postperformance assessments and is cost-and time-ineffective (Hanin, 2000; Raglin & Hanin, 2000). Finally, the direct method, if used without recall of individually best and worst performances, has a very limited and sometimes dubious value in prediction. On the other hand, it is important to explore the accuracy of zone estimation in direct assessments using different methods. Kamata, Tenenbaum, and Hanin (2002) proposed a probabilistic approach to zone estimation based on frequencies of different performance levels related to corresponding perceived or objective measures of emotion intensity. This exploratory study aimed to improve the categorical approach to zone estimation using two hypotheti cal cases with 50 and 33 data points, respectivelv and laboratory data (105 trial observations) from a single individual (Freeman, 1940). The relationships between reaction time (performance) and palmar skin resistance ■„aA Tn determine the IZOrs and their (anxiety) were examined, lo aetei mi CJI associated probabilistic curve thresholds observable pe, formance outcomes were categorized into four levels (poor, moderate, good, and excellent), and then intensity scores were regressed onto the corresponding performance categories using logistic ordinal regression. The regression coefficients were used to establish emotion-related probability curves associated with each performance category. Thus, for each performance category, a range of arousal/affect level was determined so that within this range the probability of performing at this level was higher than in the other performance categories. It was also revealed that the probabilistic method of zone estimation had wider zones than in the traditional method of estimation. Additionally, more correct classifications within the zones and fewer incorrect classifications outside the zones were obtained. These findings and the subsequent replication studies (Cohen, Tenenbaum, & English, in press; Golden, Tenenbaum, & Kamata, 2004) provide preliminary evidence of how to improve the accuracy of categorical assessments of performance-related emotion zones. These results should be accepted with caution, however, because the Kamata et al. (2002) study used only hypothetical and laboratory data. Again, many questions arise: What is the minimum number of observations in each performance category required to estimate the probability-based zones? This method of estimation of intensitv ™„„ number of direct observations lei11" " ^ and ineffective in terms ToT* * ^ lmPractical - few observat^^^ *« other hand, (optimal or poor) preclude rtT performa^e category the optimal or dysfunctional6 P°SSibUity of establishing (optimal or poor) prec.ude the posXh the optrmal or dysfunctional zone ŕ curves be used to predict f t probability point, there is no empirical Perf°rmance? At this probability curves for differ eVldenCe to suggest that the on observations of only actuľ^^01""1311^ level« based f0hmanCe Can P^dict futúr per ľ' b£St ever> Per" observations using frc^^^-Cl^ificationof Í* See*s circular Co / 0bs«ved performance a soľ Pe;f°rmanCe * eaľh f""8 * variability of Hoľirtry (and ;he probabiiit>- emotion effP * he Pr°babilitv y C ran§es> be esti--P'-ationso V" Perf-manCe? ^ to direct 0ttheseprobability.ba^hat are the practical yba-d zones. This method , present form has a strong categorical focus ,„ , u ^ndthe zones of intensity is sti,l not ass^^VT ' other questions prov.de directions for futlJre J^e lore radical approach involving the estimation of el/-impact on performance along the entire range of worki™ intensity ("mtenstty-tmpact contingencies; Hanin 997 2000: Robazza, Bortoli, & Hanin, in prcss) . ' ^ described later in the chapter. ,er|y from Anxiety to Multiple Emotions There is a growing consensus in applied sport psychology that prediction of athletic performance should be based on multiple pleasant (positively toned) and unpleasant (nega tively toned) emotions rather than only on precompetition anxiety (Cerin et al., 2000; Crocker et al., 2002; Gould & Tuffey, 1996; Hanin, 1993, 1997, 2004; Jones & Hanton, 2001; Kerr, 1997; Lane, Terry, & Karageorghis, 1995; Lazarus, 1993; Raglin & Hanin, 2000; Robazza, 2006)' Substantial empirical evidence indicates that unpleasant emotions do not always harm athletic performance. For instance, such emotions as anger, anxiety, and tension can sometimes be beneficial in competition (for reviews, see Hanin, 1978, 1995; Jones, 1995; Raglin, 1992; Raglin & Hanin, 2000). These findings are in accord with the earlier observations and anecdotal evidence indicating that highly skilled and experienced athletes can deliberately use relatively high anxiety to their advantage (Hanin, 1978; Mahoney & Avener, 1977). As a result, these expert performers often perceive anxiety as facilitating their performance (Jones, 1995). On the other hand, the findings indicate that pleasant emotions are not always beneficial for successful performance (Hanin, 1997, 2000). Too much of some pleasant emotions can sometimes lead to a poor performance due to complacency and underestimation of task demands and insufficient focus and dysfunctional energy levels (too high °rtoo low). Therefore, although some athletes perform up t0 their potential when they are stress-free, others deliberately generate and use competitive stress to their advance as an additional resource and a tool for mobil.zat.on in urgency situations. nmn M«ch of the earlier research proceeded from a nomo-**c perspective with the aim of making predictions ;«ng athletes and exercise participants in g 1996; Vallerand, 1997). Recent numerous stud ho*ever, have begun to reflect an idiograph.c pe spe ^ the aim of making predictions about: Ubsets of athletes (Hanin, 1995, 1997. 2000, 2004, a' 2°06; Vallerand & Blanchard, 2000). Emotion! in Sport 43 Although precompetition anxiety is an important stress-related emotion, it \s still only pan of the emotional mix that influences athletic performance. Determining (he interactive effects of emotions enhancing and impairing sporting activity is crucial for an accurate prediction of emotion-performance relationships. In (his case, a high probability of individually successful performance is expected when combined maximum enhancing and minimum impairing effects are observed. On the other hand, a high probability of individually average and poor performance is expected when a combination of high enhancing and high impairing effects or low enhancing and low inhibitory effects are observed. Finally, a high probability of poor performance is expected when low enhancing and high inhibitory effects are observed. In the case of pleasant-unpleasant and optimal-dysfunctional emotion intensities, it is important to assess interactive effects of four different categories of emotions: P+ (pleasant optimal), N+ (unpleasant optimal), P-(pleasant dysfunctional), and N- (unpleasant dysfunctional). Therefore, the IZOF principle was further developed to account for these interactive effects. With the development of individualized emotion profiling (Hanin, 1997, 2000; Hanin & Syrja, 1995, 1996), the extended IZOF concept is used to describe separate and interactive effects of both pleasant and unpleasant emotions using athlete-generated items. Specifically, the individual zone of optimal intensity is identified for each functionally optimal emotion, and the individual zone of dysfunctional intensity is identified for each dysfunctional emotion. In both cases, recall is used to examine past performance history rather than wait and see when successful and extremely poor performances occur. Past experiences were used to predict present and future performances. It is assumed that there are IZOFs in some emotions (P+, N+) within which the probability of successful performance is the highest. There are also dysfunctional zones in other emotions (P-, N-) within which the probability of poor performance is the highest. Optimal and dysfunctional intensity levels can be low, moderate, or high and vary for the same and different emotions in different athletes (Hanin & Syrja, 1995). Moreover, it is possible to estimate functionally optimal and dysfunctional effects, separately and jointly, only when these emotions are near or within these previously established individual zones. In other words, the total effect of pleasant and unpleasant emotions on performance appears to be determined by the interaction of optimal and dysfunctional effects. Although functionally optimal emotions are important predictors of 44 Motivation, Emotion, and Psychophysiology Harmful effects (N-P-) Emotion effects High Low High Average performance Successful performance Helpful effects (P+N+) Low Poor performance Average performance Therefore, previously established poor performances. Thererorc, r J ... »f„i ^ individualized criteria to predict indi-zones were useful as inaiviuu Figure 2.3 Interactive effects of enhancing and harmful emotions. successful performance, they alone may not be sufficient due to the fact that emotional experiences involve mixed feelings. Therefore, potential detrimental effects of dysfunctional emotions should be considered as these emotions are sometimes experienced at the same time as optimal emotions. Four quadrants in Figure 2.3 illustrate this principle in a matrix form, and the IZOF iceberg, or bell-shaped emotion profile, visually represents interactive effects (Hanin, 1997, 2000, 2003). Therefore, the notion of a zone, as applied to a wide range of pleasant and unpleasant emotions, seems appropriate in providing individualized criteria to evaluate both optimal and dysfunctional effects separately and jointly. Empirical research revealed a high degree of interindi-vidual variability in the intensity and content of idiosyncratic optimal and dysfunctional emotions related to individually successful and poor performances. It was also shown that different athletes perform up to their potential experiencing emotions of different content and intensity, and there is no universal intensity level and zone that are similar and optimal or dysfunctional for all athletes. Beyond Optimal Intensity Zones Prediction of individual performance based on contrasts of precompetition emotional states with previously established IZOFs in multiple emotions received fairly good empirical support (Annesi, 1998; Hanin, 2000, 2004; Robazza, Pelizzari, & Hanin, 2004). In most cases, the optimal and dysfunctional zones were established using the focused recall procedures described earlier (Hanin, 2000 2003; Hanin & Syrja, 1995, 1996). This proved to be' effective with highly skilled and experienced athletes, who are usually well aware of their personally significant experiences, and meta-experiences, related to successful and zones were vidual performance. In earlier research on optimal anxtety, the mam emphasis was on personally best and worst performance and emo-these two personally significant ircn uii — - » — illy best and worst performance and emotions accompanying these two personally significant situations. However, it was not known ,f experienced emotions represented also an optimal or dysfunctional (repeated) pattern All other performance levels were assumed to be between these two extremities. When the focus of research shifted from anxiety to pleasant and unpleasant emotions, a new construct was proposed: a notion of individual performance range with distinctions between personal best and personal worst categories, including personally standard and substandard performances. Although the initial approach was based on categorical assessments (in or out of the zone), a more comprehensive approach (Hanin, 1997, 2000) required continuous (along the entire working range of intensity) estimation of what was beyond the zones of intensity and performance ranges. Such an assessment strategy is important when multiple items of emotion and nonemotion experiences are used to estimate the partial and total impact of emotional experiences on performance. The IZOF-based research in performance anxiety has also indicated that if intensity was closely out of the zones, performance deteriorated, but in some cases, performance did not deteriorate when intensities were further from the zones (Turner & Raglin, 1996). Finally, several IZOF emo-tion studies revealed that differs.** , . . . nat mtrerent emotions can be optimal or dysfunctional or both Th»o f j-each emotion (in each JT^ZL^ *" have a different effect (op m ? Categ°neS) ^ ing on its intensity J^T"*^"10™1^^ approach, as a practical tool for r 8 Cate§orical and dysfunctional intensity i estimate of optimal we have a total intensity scorePRCtlCally aCCePtable when emotion on performance when th ^ the impaCt °f the zones or even along the e f lntensity is well beyond ty? To answer this question & ^ °f WOrkin§ intens-establishing intensity-irnpact ' * sinuous approach in tion (partial effect) and for a ™mn&nc^ for each emo-needed. The multiple emotion m°UOnS (t°tal was assessment of part]alef l°assessment requires the goncally and the uSe of ^^"T^ rather than cate- in an explorato skiers estimated pero. 12 toP Finnish emotion on their VCd effects of f ^-country the,r Performance aLJ Self-generated g tne entire range of Emotions in Sport 45 intcnsity (Hanin, 1997, 2000). As a result, the intensity. hupac, contingency for each idiosyncratic emotion generated by the athletes was created. This study provided initial empirical support for a more detailed estimation of the interactive effects of different emotions on athletic performance, Specifically, it was shown that being outside the optimal /ones may indeed produce a less enhancing effect, or even have a detrimental effect (e.g., an absence of motivation or energy), on individual performance. Similarly, being out of the dysfunctional zones in performance-inhibiting emotions can be not only less detrimental but sometimes can even enhance individual performance effects (e.g., an absence of fatigue or depression). Therefore, a more accurate estimation of total emotion impact on performance was possible, providing it was based on individualized intensity-impact contingencies developed by athletes for each emotion. The development of intensity-impact contingencies is based on an athlete's awareness and ability to report his or her own experiences. Additional research is needed to estimate how accurately athletes of varying skill and experience are able to do such estimations and how accurate are the predictions that are based on these contingencies. A recent study by Robazza et al. (in press) examined the perceived effect of idiosyncratic emotions and bodily symptoms on athletic performance along the entire emo-tion-intensit\ range. The participants were 35 elite Italian athletes (16 females and 19 males) competing in either figure skating or gymnastics. Idiosyncratic emotional descriptors were rated on Borg's Category Ratio (CR-10) scale to estimate the perceived impact on performance and hedonic tone for each level of emotion intensity range. The findings revealed large interindividual variability in the content of emotions as well as in the shape of the curves representing the intensity-impact contingencies. At the group level, the emotion-performance link was positively linear for optimal-pleasant emotions, bell-shaped lor optimal-unpleasant emotions, and negatively linear for both dysfunctional-unpleasant and dysfunctional-pleasant emotions. Future research should focus on how intensity-impact contingencies can be used in the estimation of total impact to predict individual performance. By definition, emotion is an unfolding process (Folkman & Lazarus, 1985). Its dynamics involve two basic dimensions: context and time (Hanin. 1997). The context dimension is an environmental characteristic reflecting the impact of situational, interpersonal, intragroup, and organizational factors on emotion intensity and content in sport sellings. Emotional experiences of varying form, content. and intensity are usually observed in different settings (context). Situational impact is manifested in emotions experienced In practices and competitions during athletes' anticipated or real contacts and interactions with significant others (a partner, a coach, and teammates). Context dimension also includes culturally coded and culturally determined beliefs of participants about the expected impact of specific emotions on their performance and about the rules of emotion display (expression or suppression) in a particular subculture. Current emotion research in sport psychology focuses on several contexts, such as successful and unsuccessful competitions of varying significance (local, national, international), and different practices. Additionally, there are a number of individually difficult situations, or specific performance episodes, that have a special meaning for athletes and teams (weather conditions, competition sites, good and bad memories of past performances). These situations may also include qualifications, performance in the finals, play-offs, meeting a weaker opponent, and performing after repeated success or a series of slumps. As for the time dimension, traditionally it is associated with a short-term situational emotion dynamics across three interrelated situations: prior to an action, during task execution, and after performance in a single competition (or practice; Cerin et al., 2000; Hanin. 1993, 1997, 2000; Jones, 1991; Syrjä, Hanin, & Pesonen, 1995). The time dimension, however, is not limited to what is going on cross-sectionally in a single competition. Moreover, cross-sectional assessments do not usually reflect the specifics of transitions of emotional experience from pre-event to midevent to postevent situations (Hanin & Stambulova, 2002; Ruiz & Hanin, 2004a). Thus, to reflect a real dynamics of emotional experience as a process, cross-sectional measures should be supplemented by qualitative methods, such as narratives or video-assisted self-confrontation interviews (Hanin, 2003; Sěve, Ria, Poizat, Saury, & Durand, in press). Long-term temporal dynamics are related to emotion-performance relationships during a competitive season (seasons), the 4-year Olympic cycle, or an athlete's sports career. The best indicators of long-term development of emotional experiences are relatively stable emotion patterns and especially meta-experiences. In the assessment of temporal patterns of emotional experiences, future researchers should include both topological (phases, cycles, sequencing, periodicity, timing) and metric (duration, frequency) characteristics. Research on topological characteristics of temporal patterns in the dynamics of emotions in sport remains nonexistent. 46 Motivation, Emotion, and Psychophysiology Finally emotion-performance relationships are dynamic and bidirectional: pre-event emotions produce beneficial or detrimental effects on performance and ongoing performance process (successful or unsuccessful) affects an athlete's emotional state. Thus, to describe emotion-performance relationships, it is important to establish the patterns of emotion impact on performance and performance impact on emotions. This latter aspect of performance-emotion relationships is especially important in research into temporal patterns of emotions across several game episodes, especially in ball games and combat sports (Seve et al., in press). Most sport events are continuous, and in long duration sports, much happens between the start and the finish. Therefore, temporal patterns are important to consider in explaining how emotion affects performance and performance-induced emotions. For instance, preperformance situations can be explained by the "anticipated gain-loss" appraisals involving challenge and threat and related emotions (Lazarus, 2000). However, what happens when "occurred gain-loss" appraisals involving benefit and harm are triggered? And how do intermediate occurrences during performance affect appraisals and emotional experiences? All these are promising directions for future researchers. EMOTION-PERFORMANCE RELATIONSHIPS A detailed description of defining characteristics of emotional experiences based on systematic observations of athletic performance is an important starting point. However, to explain emotion-performance relationships in sport, it is also necessary to look at the antecedents and consequences (effects) of emotions relating to athletic performance. After that, a tentative explanation of individual differences in emotion response is possible. In this section a brief overview of antecedents and consequences of emotional experiences and two interconnected explanations of their effects on performance are suggested. Antecedents of Emotions in Sport According to Vallerand and Blanchard (2000), theory and research on antecedents of emotions deal with psychological processes eliciting emotions with the aim to understand and predict how an individual will feel in a given sport situation. Several existing cognitive theories and research on antecedents of emotion in sport illustrate well past research and recent trends potentially important in sport settings Vallerand and Blanchard provide a detailed revlew of the early contributions to theory on emotion, selected appraisal theories, goal and motivational theory, and research Readers are also referred to another excellent review of selected cognitive theories and sports-specific models by Crocker et al. (2002) that deals with emotion antecedents. Most of these approaches emphasize the role of a variety of intrapersonal determinants of self-directed emotions including individual differences in traitlike characteristics. These are achievement needs, anxiety, mastery orientation, cognitions (expectancy of success), efficacy beliefs, causal ascriptions, and incentives related to goal orientations and their sources or locus (Hareli & Weiner, 2002). Weiner's extension of his previous attribution-emotion model suggests that interpersonal context gives rise to a variety of socially related emotions and personality inferences that have far-reaching consequences. Specifically, a good deal of individuals' self-definition and emotional experiences are derived from how they are perceived and the feelings they elicit from others in achievement settings (p. 183). For instance, just as the player is experiencing different emotions based on the task outcome and the perceived cause of the outcome, involved observers (teammates, coach, fans) also are experiencing different emotions. Self-directed emotions include pride, gratitude, shame and guilt, and hopelessness; other-directed emotions are pride, envy, admiration, schadenfreude (joy at the shame of another), sympathy and contempt, anger, arrogance, modesty, and deceit soon Tan \ ^ * ^ rcsea«* *""tion in sport an emphasis on self- and other-directed social emotions. This is a neglected are-, of dlrected social eral and sport psychology.Th^r ~ ^ personal and intragroup deter^f^8 °" ^ experiences. For instance, several r emottonal personal and intragroup anxietv " ^ StUdies of inter' ples of how emotions can reflect" S nerf^ > -perf0nnanee JOCused ects _ -aPProaches f ' J°nnc {nce outcome.'' assess anxiety effectsdire^„ ^ sequence of anx-nance outcomes. °n the anxiety "onships did not Emotions in Sport 49 Research provides reasonable empirical support for the validity and potential utility of the direction construct in thc assessment of situational states and relatively stable patterns of anxiety. However, it should be recognized that optimal and dysfunctional effects of high and low anxiety on athletic performance are well-known in competitive and especially in elite sports (Hanin, 1978, 1986, 1995; Mahoney & Avenir, 1977; Raglin, 1992). Moreover, it is not surprising that elite athletes sometimes experience lower anxiety intensity and rate its effects as more facilitating than do nonelite and less experienced athletes. Although directional research seems intuitively appealing, in its present form it has several limitations. First, the construct of emotion effect (direction) has been neither defined nor adequately described. Second, similar to test anxiety studies, current research is limited to rating only the extent to which anxiety is either helpful (facilitating) or harmful (debilitative) to an athlete's performance. These ratings fail to indicate the way a specific anxiety intensity affects (or does not affect) an athlete's performance process positively or negatively. Third, in most cases, researchers failed to collect performance data directly to examine anticipated and actual impact of anxiety intensity on performance (see, e.g., Jones & Hanton, 2001). Therefore, it is still not clear if athletes who rated anxiety as facilitating really succeeded and those who rated anxiety as debilitating really failed to perform up to their potential. Fourth, it is also not known if the direction ratings of similar anxiety intensity are stable over time or if they change from competition to competition. Fifth, it is not clear how direction scores, in their present form, can be used for prediction of individual performance. Finally, although the directional approach begins to consider different feeling states (pleasant and unpleasant), the anxiety-oriented framework does not estimate the functional impact on performance of a wide range of pleasant and unpleasant emotions. Two questions are relevant to the discussion of emotion functionality: Are negatively toned emotions invariably detrimental to sporting performance? Are positively toned emotions always beneficial for performance? Numerous IZOF-based studies (Hanin, 1978, 1986, 1995, 1997, 2000; Hanin & Syrja, 1995, 1996; Jokela & Hanin, 1999; Raglin & Hanin, 2000; Robazza, 2006; Ruiz, 2004; Ruiz & Hanin, 2004a, 2004b; Syrja, 2000) provide strong empirical evidence suggesting a clearly negative response to both questions. In other words, unpleasant emotions can somettmes be helpful for performance (see Hanin, 1978, 1986; Hardy, 1990; Jones, 1995; Jones & Hanton, 2001; Ruiz, 2004), and pleasant emotions are sometimes harmful for performance (sec Carver, 2003; Fredrickson, 2001; Fredrickson & Losa-da, 2005; Hanin, 1993, 1997, 2000). Thus, the view that emotion valence is the only or a major predictor of the effect of emotion or its regulation is oversimplistic at best (Cole, Martin, & Dennis, 2004). Therefore, attempts to propose the notion of positive and negative anxiety based on its perceived effects seem questionable at best. Much confusion in this positive-negative anxiety debate (Burton & Naylor, 1997; Hardy, 1997; Jones & Hanton, 2001) comes from a failure to distinguish between emotion content, emotion intensity, and emotion functionality (helpful or harmful effects). For instance, Jones and Hanton argue that anxiety by definition is a negative (unpleasant) feeling state but claim that the CSAI-2 does not measure competitive anxiety directly, but only the symptoms associated with the response. They believe that "if a negative score on the direction scale is revealed then this signifies a state of anxiety. If a positive direction score is found, this points to another state previously mislabeled as anxiety" (p. 393). This assumption is actually true if it suggests that there are mixed emotions, besides pure anxiety, that add to positive impact on performance. However, this assumption is not true, and is even contradictory, if labeling of anxiety state depends entirely on a negative direction score. Qualitatively, anxiety is a negatively toned unpleasant state reflected in several specific symptoms (feelings of tension, apprehension, nervousness, etc.). Actually, anxiety and nonanxiety labels describe fixed or conventionally defined emotion content, whereas functional effects represent a different characteristic. Thus, using an athlete's own vocabulary of emotion labels along with researcher-generated items could be instrumental in the partial solution of this problem. The main issue in emotion research now is not only to rate the perceived impact of emotions, but to identify, for instance, in what way high, moderate, or low anxiety (or any other emotion) is helpful or harmful to athletic performance. Hanin and coworkers (Hanin & Syrja, 1995; Syrja, 2000) collected qualitative data describing how highly skilled ice hockey and soccer players perceive the functional effects of facilitating and debilitating emotions for their performance. Two major functions emerged in the content analysis of players' interpretations of perceived emotion effects: enhancing or detrimental to effort and skill. For instance, a player who experiences 50 Motivation, Emotion, and Psychophysiology dissatisfaction perceives it as a helpful emotion because this emotion helps him or her to try harder, to maintain a fighting spirit, to be better than his or her opponent, to put more effort into the game, and to be more alert. Harmful effects of too much satisfaction (complacency) are reflected in being too concerned with success, not trying to play better, being too arrogant, not careful, and too risky; as result, skating becomes difficult (Hanin & Syrja, 1995, pp. 180-181). A more detailed description of perceived functional effects of selected emotions across four global categories (P+, N+, P-, and N-) is found elsewhere (Ruiz, 2004; Syrja, 2000). Explaining Individual Differences Numerous empirical studies revealed large interindividual variability of emotion intensity and emotion content in athletes performing similar and different sporting tasks. How can these findings be explained? Why do some athletes perform well while experiencing high anxiety, whereas others fail to cope with competitive stress? Why is emotion content different in different athletes performing the same task? I propose two possible explanations to account for these differences: a resource-matching hypothesis, based on the construct of internal and external resources, and two constructs, energy mobilization and energy utilization (Hanin, 1997, 2000, 2004). The construct of internal and external resources proposed here is not new. For example, it is used in the conservation of resources (COR) model proposed by Hobfoll (1989) to define and explain psychological stress. Examples of broadly defined resources include not only personal characteristics (self-esteem, mastery, and well-being) but also interpersonal, material, and work-related resources. The basic tenet of the COR model is that people strive to retain, protect, and build resources because the potential or actual loss of these resources is a threat and a source of psychological stress. From this perspective, psychological stress is defined as a reaction to the environment in which there is (a) the threat of a net loss of resources, (b) the net loss of resources, or (c) a lack of resource gain following the investment of resources. There is a clear overlap of these ideas with the relational themes and appraisal patterns (anticipated and occurred) proposed by Lazarus (2000). Hobfoll also proposed an instrument to measure a gain and a loss of resources that was used in empirical studies with different populations outside the sport setting. The life span model of developmental challenge proposed by Hendry and Kloep (2002) employs the constructs of resources and challenges to explain the processes of human growth. Examples of potential resources include , ■ , j- v„„c chpalth personality, talents, mte-biological dispositions (heann, p ° .:..^aci:V mrm resources Itrnct •cat aisposiuvuj y*----- * ice, body shape, attractiveness); social resources (trust, attachment, size and quality of network); skUls^ (baste, learning, social, psychomotor); self-efficiency (self-efftca-cy appraisals, experience with success, assurance from others, locus of control); and structural resources (country, race, class, family, income, gender). To explain intraindividual and interindividual variability of emotion content and intensity in similar and different performance situations, a resource-matching hypothesis was proposed (Hanin, 2000, 2004; Hanin & Stambulova, 2002, 2004). Based on the idea that emotional experiences reflect person-environment interaction, it was suggested that it is not so much the task requirements per se that determine optimal and dysfunctional content and intensity of situational emotional experiences but an interaction (match or mismatch) between task demands and an athlete's resources (available, recruited, and utilized). In competitive sport, resources are defined as psy-chobiosocial assets that determine athletes' ability to perform consistently up to their potential. Here the emphasis is on how available resources are identified and then systematically and effectively recruited, used, recuperated, and further developed. Thus, for instance, a complex task can be very easy for an athlete with sufficient resources that can be recruited when needed and utilized effectively. In contrast, a task generally considered relatively easy can be very demanding and difficult if an athlete is unable to recruit available resources or not ready to use them efficiently (Hanin 2003 9004- u • n 2004 Ruiz *h .I ' Hamn & Stambulova, 2002. ' Kuiz & Hamn, 2004a, 2004b) The resource-matching hv™n,» • potential causes of intramdiidu^tdS1Sf ^ ability in optimal emot.on content'T^^*1 Vari" include interindividual diff density. These resources, (b) the ability to recri^l * 00 available and place, and (c) the skill 3t the ri8ht time there are clear imraindtvidu"? efficient^- F^Hy, ences in situational reading lnterindividual differ- perate these resources rCCruit' utilize, and recu- The four categories of ■ZOF mode, and ™o,ion conten, d •or (hedonic ,one ™m *= ™eracti„„ of two fac_ reflect a resoi,™ , Pert°rniance f,,^ «°» =nd ovde aSed '"'"PreraUon " "0ni"i,>') a'S° Purnal pleasant emotions ?-y' pre§ame ^„Vy-eo„.fl .. descriptions in psychology]" Lerij J° V Psikhohgii [Systems versity Press. n§rad. Russia: Leningrad Uni- Golden, A. S., Tenenbaum, Q & v related performance ^ta, A. (2004). Affect- ing affect to performance. Inter^T* meth°d *>r Exercise Psychology, 2, ^ ^1 of Sport and Gould, D., & Tuffey, S. (1996 research: A review and Cr»; S of optimal f 9(1), 53-68. Ul(lue- Anxiety St Phoning Hackfort, D., & Schwenkrn ^ Singer, M. Murphey, & V^J' P- «993) An . research on sport ' Tenr>ant rj ^ In R N- Macmi.lan. "**«**.v (pp * tent and functional mn (2°°3)- density ■ , C'enCe-in athletes. Journal oJT^^^ZT^ C°n' p rts Science^ 2l emotions revolutionary synthesis. Emotions in Sport 57 , c.. Bortoli, L., & Hanin, Y (2004.\ r>ra Kl1bn/Ai- ^' ' iw). rre-competition .motions, bodily symptoms, and task-specific qualities as Iiredictors of performance in high-level karate athletes Jour-mil of APPIied SP°rt Psychology, 76(2), 151-165. roMIM, C, Bortoli, L., & Hanin, Y. (in press). Perceived effect 0f emotion intensity on athletic performance: A contingency-based individualized approach. Research Quarterly for Exercise and Sport. Robazza, C, Pellizzari, M., & Hanin, Y. (2004). Emotion self-regulation and athletic performance: An application of the IZOF model. Psychology of Sport and Exercise. 5(4), 379_404. Ruiz, M. C. (2004). Anger and optimal performance in karate: An application of the IZOF model (University of Jyvaskyla, Jyvaskyla, Jyvaskyla Studies in Education). Psychology and Social Research, 254, 130. Ruiz, M. C, & Hanin, Y. L. (2004a). Idiosyncratic description of anger states in skilled Spanish karate athletes: An application of the IZOF model. Revista de Psicologia del Deporte, 25(1), 75-93. Ruiz, M. C, & Hanin, Y. L. (2004b). Metaphoric description and individualized emotion profiling of performance related states in high-level karate athletes. Journal of Applied Sport Psychology, 16(3), 1-16. Russell, J. A. (1980). A circumplex model of affect. Journal of Personality and Social Psychology, 39, 1161-1178. Russell, J. A., Weiss, A., & Mendelsohn, G. A. (1989). Affect grid: A single-item scale of pleasure and arousal. Journal of Personality and Social Psychology, 57(3), 493-502. Schimmack, U. (2001). Pleasure, displeasure, and mixed feelings: Are semantic opposites mutually exclusive? Cognition and Emotion, 75(1), 81-97. Schlosberg, H. (1954). Three dimensions of emotion. Psychological Review, 61(2), 81-88. Schwartz, R. M. (1997). Consider the simple screw: Cognitive science, quality improvement, and psychotherapy. Journal of Consulting and Clinical Psychology, 65, 970-983. Seve, C, Ria, L., Poizat, G., Saury, J., & Durand, M. (in press). Performance-induced emotions: Experienced during high-stakes table tennis matches. Psychology of Sport and Exercise. Simon, H. A. (1967). Motivational and emotional controls of cognition. Psychology Review, 74, 29-39. Smith, R. E. (1996). Performance anxiety, cognitive interference, and concentration enhancement strategies in sports. In I. G. Sarason, G. R. Pierce, & B. R. Sarason (Eds.), Cognitive interference: Theories, methods, and findings (pp. 261-284). Hillsdale, NJ: Erlbaum. Smith, R. E„ Smoll, F. L., & Schutz, R. W. (1990). Measurement and correlates of sport specific cognitive and somatic trait anxiety: The Sport Anxiety Scale. Anxiety Research, 2, 263-280. Spielberger, C. D., Gorsuch, R. L., & Lushene, R. E. (1970). Manual for the State-Trait Anxiety Inventory (STA1). Palo Alto, CA: Consulting Psychologist Press. Stambulova, N. B. (2000). Athletes' crises: A developmental perspective. International Journal of Sport Psychology, 37(4), 584-601. Syrja, P. (2000). Performance-related emotions in highly skilled soccer players: A longitudinal study based on the IZOF model— Studies in sport, physical education and health (Vol. 67). Jyvaskyla, Finland: Jyvaskyla University Printing House. Available from http://selene.lib.jyu.fi:8080/gradu/g/psyrjan .pdf. Syrjä, P., & Hanin, Y. (1997). Measurement of emotion in sport: A comparison of individualized and normative scales. In R. Lidor & M. Bar-Eli (Eds.), Proceeding of the 1SSP IX World Congress of Sport Psychology (Pt. 2, pp. 682-684). Netanya, Israel: Wingate Institute. Syrjä, P., & Hanin, Y. (1998). Individualized and group-oriented measures of emotion in sport: A comparative study. Journal of Sports Sciences, 16(5), 398-399. Syrjä, P., Hanin, Y, & Pesonen, T. (1995). Emotion and performance relationship in soccer players. In R. Vanfraechem-Raway & Y. Vanden Auweele (Eds.), Proceedings of the 9th European Congress on Sport Psychology (Pt. 1, pp. 191-197). Brussels, Belgium: Belgian Federation of Sport Psychology. Thompson, R. A. (1994). Emotion regulation: In search of definition. Monographs of the Society for Research in Child Development, 59(2/3, Serial No. 240), 25-52. Turner, P. E., & Raglin, J. S. (1996). Variability in precompeti-tion anxiety and performance in college track and field athletes. Medicine and Science in Sports and Exercise, 28(3), 378-385. Vallerand, R. J. (1997). Towards a hierarchical model of intrinsic and extrinsic motivation. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 29, pp. 271-360). San Diego, CA: Academic Press. Vallerand, R. J., & Blanchard, C. M. (2000). The study of emotion in sport and exercise. In Y. L. Hanin (Ed.), Emotions in sport (pp. 3-37). Champaign, IL: Human Kinetics. Vygotsky, L. S. (1984). Sobranie sochinenii [A 6-volume collection of works]. In D. B. El'konin (Ed.), Child psychology (Vol. 4, p. 247). Moscow, Russia: Pedagogika. (Original work published 1926) Wang, Y, & Morgan, W. P. (1987). Convergent validity of a body awareness scale. Medicine and Science in Sports and Exercise, 19, S579. Watson, D., & Tellegen, A. (1985). Towards a consensual structure of mood. Psychological Bulletin, 98, 219-235. Weinberg, R. S. (1978). The effects of success and failure on patterning of neuromuscular energy. Journal of Motor Behavior. 10, 53-61. 58 Motivation, Emotion, and Psychophysiology Weinberg, R. S., & Hunt, V. V. (1976). The interrelationships between anxiety, motor performance andelectromiography. Journal of Motor Behavior, 8, 219-224. Woodman, T., & Hardy, L. (2001). Stress and anxiety. In R. N. Singer, H. A. Hausenblas, & C. M. Janelle (Eds.), Handbook of sport psychology (2nd ed., pp. 290-318). New York: Wiley. Wurth S & Hanin, Y. L. (2005). Der emotionale verstarb, tand im Zusammenhang mit sportververletzungen [Pre.per. formance emotional state related to sport mjunes]. Leipzigr Sportwissenschaftliche Beitrage (LSB), 46(1), 117-143. Yerkes, R. M., & Dodson, J. D. (1908). The relation of s •j:^.. „t hahit-fnrmatinn. Journal trengih rkes R M., & uou.wh, j. ^. v----— -«v„ of s'timulus to rapidity of habit-formation. Journal of c parative Neurology and Psychology, 18, 459-482. orti-