The Role of Personality in Sport: Conceptual and Methodological Challenges Eugene V. Aidman University of Adelaide, Australia The Science of Personality we are: – different from anyone else (uniqueness) – remain ourselves across situations (consistency) These differences are measurable Thurstone’s law: if something exists, it exists in some amount and can therefore be measured Personality research: study of measurable individual differences – but what are they? Situation-free dispositions (i.e. aggregated across time) vs situationally hedged dispositions = conditional and interactive with the situations in which they are expressed (Mischel, 2004) Personality & Sport Compared to non-sport playing controls on 16PF, national level competitors are (Williams, 1985): – – – – higher emotional stability greater mental toughness more self-assured more trusting Getting into an Olympic squad in wrestling (Silva et al., 1985) linked to (16PF) sociability, boldness, emotional stability and apprehension Mood States and Performance Morgan & Hammer (1974) - Terry (2000) better performing athletes display more positive mental states: – less anxious – less depressed – less fatigued – less confused – more vigorous (and extroverted) Mental health profile Positive Mental Health Profile: (Morgan & Johnson, 1978) found lower levels of psychopathology (MMPI) in more successful University oarsmen However: hardly any replication – e.g. Brown, Morgan & Kihlstrom (1989) found no significant associations between MMPI profiles of collegiate athletes and their athletic success Anxiety and Performance Levels - high vs low - are insufficient state - trait anxiety (Spielberger) cognitive appraisal of threat: – facilitative anxiety: stress response as excitement – debilitative anxiety: stress response as threatening Personality & Achievement Davis & Mogk (1994) compared elite, sub-elite, non-elite and non-athletes on EPQ, Sensation-seeking and Achievieng Tendency scales: – the key factors linked to the level of competitive achievment: • emotional stability • and achievment motivation Personality and success Piedmont, Hill & Blanco (1999): coach ratings of performance and game stats linked to the Big Five profiles of elite soccer players: – Neuroticism / emotional stability – Conscientiousness / «will to achieve» – acceptance of criticism: «coachability», in turn linked to higher self-esteem Personality and Performance Origins in Org- and Ed- psychology: selecting for success Personality-Related Position Requirements Form (PPRF; Raymark & Schmidt, 1997): – based on the Big Five model (McRae & Costa, 1992) – found personality factors predictive of job performance based on specific competencies (job needs analysis) Sport Psychology is yet to follow PPRF’s lead Personality and Sport Performance sceptical vs credulos debate (Morgan, 1980) – Personality is a weak predictor of Sport Performance – but it is a Predictor Weak theory - wrong place to look for connections Weak method - hopeless in catching a connection even if there was one (insufficient design) The connection is unlikely to be DIRECT and IMMEDIATE The Role of Personality in Sport & Exercise in the long run: converting ability into achievement from promice to delivery – sub-elite to elite sport transition «here and now»: moderating the effects of circumstances on performance • • • • stress tolerance -vs- anxiety volatile motivated -vs- slack: e.g. winning from behind focused -vs- all over the place injury pronene - hardy Example 1: Personality in Long Term Achievement Elite Juniors’ transition to Senior AFL (Aidman, 2004) Method 32 elite junior players from a leading Australian Football League (AFL) club: mean age 17.8 (1.1) players profiled with Cattell’s 16PF (Form A) at the peak of their junior playing career – immediately after the season where they won the National Championship in their age group. Head Coach rated players’ performance and physical potential (5-point Likert scales) 7-year follow-up: has the player made it to senior AFL(drafted+played at least one season) or not ? Results 13 players made it into senior AFL competition 19 others ended up playing minor leagues or dropped out of the game altogether MANOVA showed no significant differences between these two groups of players on primary personality factor profiles when the players’ physical potential rated by their junior head coach was controlled for in an MANCOVA, the differences between the groups became highly significant: both on multivariate estimates (F (16, 14) = 3.506; p = .012) and on a number of individual factors Results: Group Differences ELITE JUNIOR AFL PLAYER PROFILE 10 9 8 stens 7 6 Mean 5 4 3 2 1 A B C E F G H I L M N O Q1 Q2 Q3 Q4 16PF primary factors Results: Group Differences G Q1 Q2 Q4 did not make it to senior AFL Mean SD 11.74 2.86 10.16 3.72 8.84 2.03 12.53 5.65 Successful transition to senior AFL Mean SD 13.38 3.20 8.00 3.58 6.92 2.29 9.23 2.71 MANCOVA differences p .08 .018 .003 .048 Personality in Long Term AFL Success: Elite Juniors’ transition to Senior AFL Coach Ratings ONLY: Compare with flipping a coin 16 Personality Factors Profile ONLY 16 Personality Factors Profile + ONE Coach Rating (physical potential): Aidman (1999, 2000) Predicting senior AFL performance from personality Prediction targets: 2. aggregate of senior achievement over the last 5 seasons (Alpha=.96) % variance explained 1. performance in junior championship at the time of testing 99 100 75 61 50 25 11 0 3. coach rating on a 5-point scale: "struggling vs cruising through senior league ranks" 1 2 Prediction targets 3 Conclusions: Confirmed the influence of Personality factors on sub-elite to elite sport transition in AFL however, this influence is – indirect – observable only in the long term Interaction with Ability: – Ability (physique in AFL) = entry ticket – Personality acts as a means of converting ability into achievement (from a promicing junior to an accomplished athlete) Example 2: Personality and on-the-day performance prediction (Aidman & Beckerman, 2001) Specific personality characteristics implicated: – – – – – Emotional stability Achievement orientation Conscientiousness (e.g., discipline) Self-concept (e.g., confidence) Anxiety Method Participants: 48 Australian Rules football players (M = 21.40 years, SD = 3.11 years) who played a full season with a successful Victorian Football League (VFL) club Instruments: – – – – Five Factor Inventory (NEO-FFI; McCrae & Costa, 1992) Self-Apperception Test (SAT-2; Aidman, 1997, 1999) Self-Liking/Self-Competence Scale (SLCS; Tafarodi & Swann, 1995) Stress Appraisal Questionnaire: Threatening versus Exciting Procedure – Aggregated game statistics across a complete season – ‘Credits’ score representing the effort and quality of performance for each player in every game Results Three distinct groups of players identified: groups were found to be predictably different on: – Self-discipline – Achievement Striving – Neuroticism (Fig. 1) mean scores – elite (senior players) – non-elite (reserves) – sub-elite (“swingers” – players who played at both levels) 20 18 16 14 12 10 player group: 8 seniors 6 sw ingers 4 reser ves neuroticism dutifulness achievement self-discipline Results: Interaction between personality and situation in the prediction of effort Three categories of games identified: – ‘Close Games” - in dispute for almost the entirety of the game – ‘Easy Wins’ - where the result was well in the team’s favour most of the way and no longer in dispute – ‘Bad Losses’ - where the team was well beaten most of the way and no longer in the contest Hierarchical Regression predicting game performance: – “easy win” games predictors: Self-discipline and Neuroticism – “close” games predictors: Neuroticism and Self-esteem – “bad losses” - no connection Table 1. Game performance (‘Credits’) (SD) Across Three Game types, by Stress Appraisal Threatene d Appraisal Low High Excitement Appraisal Close Game Easy Win Bad Loss Low 4.762 (0.811) 5.452 (0.818) 5.590 (0.823) High 4.217 (0.796) 4.383 (0.803) 4.376 (0.808) Low 3.744 (0.817) 3.950 (0.824) 2.756 (0.829) high 4.696 (0.513) 4.592 (0.517) 4.385 (0.520) Stress appraisal and game performance 6 Low threatening Low Exciting Low threatening High Exciting Low threatening High Exciting High threatening Low Exciting High threatening High Exciting 5 4 3 2 1 0 Close Game Easy Win Bad Loss Example 3: Self-esteem and Performance (Meagher & Aidman, 2004) Three aspects of Self: Cognitive: self-attributions • bright, attractive, athletic, slow etc. Affective: how we feel about these self-attributions (evaluation) • self-esteem = affective avaluation of self (Martens, 1975) Behavioural: our tendencies to behave in accordance with self-image • Self-concept as self-fulfillling prophecy: self-concept is more than self-descriptions, its a commitment to continue being oneself “as described” Rationale for Indirect Measurement of Self Global self-attitudes vs self-descriptions self-presentation distortions – – deliberate (faking, impression management) self-deceptions (genuine) affective / implicit elements of Self – – – displaced self-esteem (Cialdini, 1993) self-positivity bias (Taylor & Brown, 1988) implicit affiliation / rejection (Tesser, 1988; Suls & Wills, 1991) Indirect Measurement of Self-Attitudes: Essential Ingredients Responce latencies in mixed category discrimination tasks (IAT; Greenwald et al. 1998) (semi) projective stimulation relevant to Selfimage – fuzzy images (Ligett, 1959) / facial sketches (Aidman, 1999) replicable procedure: – semantic differential (Snider &Osgood, 1969) Relevant self-attitude scales: – – global (self-worth, self-competence) specific (ability, attractiveness, strength...) Self-reported vs indirect self-appraisal and elite swimmers’ performance (Aidman & Perry, 2000) Method: Participants 38 elite Australian swimmers (15 females and 23 males, mean age 20.1 years, SD = 2.84) participated as part of their preparation program for the 1998 World Championship Method: Instruments Self-Liking/Self-Competence Scale (Tafarodi & Swann, 1995) Cronbach’s alphas: .92 for self-liking .89 for self-competence Self-Apperception Test (SAT; Aidman, 1999) – measuring implicit self-appraisal (ISA) Cronbach’s alpha: .83-.90 for Global ISA (retest reliability 0.57 - 0.84) Method: Procedure Self-appraisal measures taken 3 months and 1 week prior to the competition (time 1) ISP (international performance ratings devised by FINA) recorded at time 1 and immediately after the competition implicit self-attitudes hypothesised to predict ISP change (positive self-affect to be associated with gains in ISP) Declared and Implicit Self-Appraisal: correlations with World ranknings (ISP) Implicit self-strength r = .33 p < .05 Declared self-liking r = .02 Ns Declared self-competence r = .04 Ns Declared and Implicit Self-Appraisal: correlations with pre-post competition change in swimmers’ ISP Implicit self-appraisal of ability (combined with its valence) r = .59 2 R =.35 p < .01 Declared self-liking r = -.2 Ns Declared self-competence r = .12 Ns Conclusions Declared self-attitudes DID NOT predict performance improvement at World Championship Implict self-appraisal of ability DID, consistent with the theoretical prediction Implict self-appraisal of strength was directly (although weakly) associated with ISP none of declared self-esteem scores were Conclusions cont’d Self-affect is conceptually and meaningfully linked to athletes’ ability to perform at their best Self-affect measurement may play an important role in predicting athletic performance at elite level But in order to fulfill this role, predictions should be (a) specific, (b) conceptually driven, and (c) matched to an adequate method of measurement (i.e. implicit rather than declared) Overall Conclusions personality effects are likely to be – Long term (e.g. converting ability into achievement) – Moderating rather than direct (e.g., moderating the effects of circumstances on performance) Situation is more than a source of noise in personality measurement – it is a key ingredient of it: “if… then…” behavioural signatures (Mischel, 2004) Types of situations with psychologically equivalent meaning (e.g., frustration) – Must be very specific – Theory driven Epilogue: behavioural signatures of aggression not an aggregate aggression score, but a profile of aggressive responding “if… then…” (Mischel, 2004) Unprovoked attacks - Aggression as an intrinsic choice Retaliatory attacks - i.e. «tooth for tooth» Frustration-driven attacks - lashing out at an obstacle escalation: mastering an aggressive response may / may not translate to its greater use Computer-game-embedded assessment (Aidman & Shmelyov, 2002) Interaction types in reverse desirability order – – – – hosts Avatar is attacked avatar Avatar’s path blocked Mimics game Avatar is allowed through Stimulus material: schematized Avatar is allowed through with a smiling greeting and extra power) facial universals (Ekman, 1999) Objectives of the game: – reach desired destination – score maximum points along the way – can be achieved through any combination of: – searching for effective expressions – searching for efficient routes – attacking the hosts – player is free to choose the tactics (may be prompted by instruction) – Avatar - player controlled expression Hosts - human-like responding to the Avatar’s expressions objective = negotiate a maze-like matrix of hosts for a reward :-) Controllable elements of expression: – mouth – eyes – eybrows each element can be made: – – – smiling neutral frowning independently of the other two Mimics measures rate of unprovoked attacks (aggression as an intrinsic choice) rate of retaliatory attacks (aggression mirroring) frustration-driven attacks (aggressive over-reaction to blockings) threatening: choosing a frowning expression intrapunitive / avoidant responding to aggression, e.g. evasion Overall - 26 measures 70 based on automated 60 standardized 50 observations psychopath 40 neurotic evangelist balanced 30 20 10 0 unprovoked retaliatory frustrationdriven Unprovoked attacks under Peaceful and Open instructions 30 Self-reported aggression (Buss-Perry total): 25 20 low high 15 10 5 0 peaceful open instruction Retaliatory attacks under Peaceful and Open instructions 70 Self-reported aggression (Buss-Perry total): 60 50 40 low high 30 20 10 0 peaceful open instruction Frustration-driven attacks under Peaceful and Open instructions 70 Self-reported aggression (Buss-Perry total): 60 50 40 low high 30 20 10 0 peaceful open instruction Correlations between Self-reported Aggression and Changes in Mimics parameters from Peaceful to Open Instruction (N=37) Bjorkvist et al., 1993 Aggression mesures Buss .& Perry, 1992 physical verbal Indirect social Indirect ratio physical verbal anger hostility Mimics measures Unprovoked attacks .06 .18 .12 .29 .05 .14 -.04 -.19 Retaliatory attacks -.17 -.03 -.08 -.04 -.01 .16 .22 -.17 Aggressive out-reaction .41** .25 .27 .49** .45** .15 .42** .29 Evasion -.22 -.16 .01 -.07 -.14 -.32* -.17 -.14 Selected References Greenwald, A.G., McGhee, D.E., & Schwartz, J.K.L. (1998). Measuring individual differences in implicit Aidman, E.V. (1999). Measuring individual differences in implicit self-concept: initial validation cognition: the Implicit Association Test. Journal of Personality and Social Psychology, 74, 1464-1480. of the self-apperception test. Personality and Individual Differences, 27, 211-228. Kihlstrom, J. (1999, September). The discovery of the Aidman, E. & Carroll, S. (2003) Implicit unconscious. Paper presented at the meeting of the Individual Differences: Relationships between Australian Psychological Society, Hobart, Tasmania. Implicit Self-Esteem, Gender Identity and Gender Meagher, B., & Aidman, E. (2004) Individual Attitudes. European Journal of Personality, 17 (1), Differences in Implicit and Declared Self-Esteem as 19-37. Aidman, E., & Shmelyov, A.(2002). Mimix: a Predictors of Response to Negative Performance symbolic conflict/cooperation simulation program, Evaluation: Validating Implicit Association Test as a with embedded protocol recording and automatic Measure of Self-Attitudes. International Journal of Testing,4 (1),19-42. psychometric assessment. Behavior Research Methods, Instruments & Computers, 34 (1), 83-89. Tafarodi, R.W., & Swann, W.B. (1995). Self-liking and selfBaumeister, R.F., (1999) Low self-esteem competence as dimensionality of global self-esteem: initial does not cause aggression. APA Monitor, 30 (1) , validation of a measure. Journal of Personality 7. Assessment, 65, 322-342. Baumeister, R.F., Heatherton, T.F., & Tice, Tallent R., & Aidman E. (1995). The impact of D.M. (1993). When ego threats lead to selfregulation failure: Negative consequences of high residential status upon quality of life in elderly women. 1995 APS Conference, abstracted: Australian Journal of self-esteem. Journal of Personality and Social Psychology, 47 (supplement), p. 119. Psychology, 64, 141-156.