The Sport Psychologist, 1994, 8, 73-86 O 1994 Human Kinetics Publishers, Inc.

Psychological and Psychomotor Skills Associated With Performance in Golf Patrick R. Thomas and Ray Over Griffith University Psychological and psychomotor skills associated with performance in golf were established through ratings provided by 165 men with golf handicaps ranging from 5 to 27. Several components of skilled performance in golf were identified through factor analysis of these ratings, followed by comparisons between lower handicap and higher handicap players. Skilled golfers (those with lower handicaps) reported greater mental preparation, a higher level of concentration when playing golf, fewer negative emotions and cognitions, greater psychomotor automaticity, and more commitment to golf. Three self-report assessment scales (measures of psychological skills and tactics, psychomotor skills, and golf involvement) were developed from the data. Contexts in which these scales can be used are discussed.

Study of the relationship between performance in sport and psychological variables has often involved comparing successful and unsuccessful performers on trait measures of personality. Responses on personality tests such as Cattell's 16 Personality Factor Questionnaire (16PF), the Minnesota Multiphasic Personality Inventory (MMPI), and the Edwards Personal Preference Schedule have been scored in a search for the "ideal profile" at different levels of competition (Silva, 1984). Although Ogilvie (1991) remains a strong advocate for the use of such test profiles in enhancing athletic performance, attempts to differentiate elite and nonelite athletes on the basis of personality traits have met with limited success (Singer, 1988). In contrast to the trait paradigm, many researchers now favor the interactionalperspectives offered by social learning theories (e.g., Bandura, 1986; Mischel, 1977). In this approach, researchers typically identify meaningful situations in sport, and then monitor individual differences in cognitive-perceptual processes in order to describe and predict the responses of athletes to those situations (Fisher, 1984). Self-report instruments developed from interactional perspectives explicitly measure psychological skills in sporting contexts. Loehr's (1982) Psychological

Pahick R. Thomas is with the Faculty of Education at Griffith University, Nathan, Australia 41 11. Ray Over is now with the Department of Psychology at La Trobe University, Bundoora, Australia 3083.

74

Thomas and Over

Performance Inventory (PPI), for example, provides a profile across seven dimensions: self-confidence, negative energy, attention control, visual and imagery control, motivational level, positive energy, and attitude control. Although Loehr offered guidelines for interpreting scores on the PPI and an illustrative example of the profile of a champion, he did not establish norms for the inventory or produce evidence for the validity or reliability of the seven scales. Mahoney, Gabriel, and Perkins (1987) used a combination of item, cluster, factor, discriminant, and regression analyses in developing their Psychological Skills Inventory for Sports (PSIS). Data analysis based on truelfalse responses provided by athletes to 51 items yielded six scales (Anxiety, Mental Preparation, Team Orientation, Concentration, Confidence, and Motivation) each of which contributed towards differentiationof elite and nonelite athletes. Mahoney et al. (1987) also developed a revised version (PSIS R-5) in which athletes use a 5-point Likert scale to rate their level of agreement with 45 items that were included in the original PSIS scale. Other investigators employing the PSIS R-5 (Greenspan, Murphy, Tammen, & Jowdy, 1989; Lesser & Murphy, 1988)confirmed that there are differences on the psychological processes measured by this test between athletes ranked from novice level to national and world-class standard. However, the status of the PSIS R-5 scale is uncertain because Tarnmen, Murphy, and Jowdy (1990) were unable to replicate the factor structure established by Mahoney et al. (1987). Instruments such as the PPI and the PSIS were designed to assess athletes across a wide variety of sports. However, the task-specific nature of psychological skills required in sport is now widely recognized. Nideffer (1985), for example, argued that the nature of attentional focus differs between sports. If such is the case, instruments intended to measure concentration and other psychological skills associated with performance in a given sport must target situations relevant to this sport. Support for the development of specific rather than general measures is provided by research demonstrating that sport-specific versions of Nideffer's (1976) Test of Attentional and Interpersonal Style (TAIS) not only have better psychometric properties than the general test, but are better predictors of sport performance (Albrecht & Feltz, 1987; Summers, Miller, & Ford, 1991; Van Schoyck, & Grasha, 1981). McCaffrey and Orlick (1989) used a series of open-ended questions to explore psychological and situational factors related to excellence in golf. They found substantial differences between touring golf professionals and club professionals. The highly skilled tour players were totally committed to golf, particularly during periods of peak performance. They set clearly defined goals, and they were more systematic in planning practice sessions and tournament play, as well as in selfevaluation after each round. Touring professionals used imagery techniques more frequently and for more facets of the game than the club professionals. Both groups developed plans for focusing attention during a round, but the touring professionals were more likely to execute their strategies throughout a tournament. In addition, they attained greater depth of concentration. Both groups acknowledged the effect of distractions on the course, particularly slow play, but the touring professionals were better able to cope with these distractions and regain a task focus. McCaffrey and Orlick (1989) concluded that maintainingcommitment, focusing, and refocusing were the skills most needed for excellence in golf. Cohn (1991) also used open-ended questions in interviewing 10 touring golf professionals, 2 club professionals, and 7 collegiate golfers. Respondents

Psychomotor Skills

75

claimed that their attentional focus while playing golf typically was narrow, taskdirected, and oriented towards the present situation rather than past or future events. Further, the execution of a skill during peak performance was perceived as effortless and automatic, requiring little if any conscious thought to control movement. At the same time respondents had the feeling of being in control of their performance and being able to regulate their emotions, thoughts, and level of arousal. Heightened arousal was generally perceived as an indicator of preparedness. Periods of peak performance were characterized by high self-confidence, with the golfers typically experiencing limited concern for the negative consequences of a poor shot. Physical and mental relaxation were also evident at times of peak performance, as was the experience of enjoyment in playing well and achieving goals. Cohn (1991) concluded that peak performance in golf is associated with positive psychological states. Success at golf thus seems to involve not only domain-specificpsychological skills, such as focusing and refocusing attention, visualization, and emotional control, but also substantial commitment of time and effort towards achieving defined goals. High levels of psychomotor competency are also required, but of a specific kind. In terms of psychomotor requirements, golf as a sport places low demands on reaction time, speed of limb movement, extent flexibility, dynamic flexibility, dynamic strength, and stamina, but requires a high level of control precision, multilimb coordination, rate control (timing), arm-hand steadiness, aiming, explosive strength, fine body coordination, and gross body coordination (see Schulz & Curnow, 1988). Researchers have acknowledged the need for valid, reliable measures of component skills underlying performance in golf and have suggested that questionnaires rather than open-ended interview techniques be developed for this purpose (Cohn, 1991; Cohn, Rotella, & Lloyd, 1990). Instead of being restricted to a specific component, such as automaticity in shot making or competitiveness, the objective should be to develop a questionnaire covering a range of component skills. The aim of the present study was to develop self-report measures of attributes associated with performance in golf. A scale, the Golf Performance Survey, designed to assess psychological skills underlying prowess at golf, psychomotor competencies, and commitment to golf is reported. Individual differences in psychological skills, psychomotor competencies, and commitment are then examined in relation to level of performance in golf. Whereas previous research has focused on players at elite levels of professional golf, factors such as correlates of success at golf are examined to see whether they apply at the amateur level and whether they distinguish skilled golfers from unskilled golfers.

Method Subjects The objective was to generate a sample of male golfers covering a range of ages and golf handicaps. The Pacific Golf Club in metropolitan Brisbane, Australia, gave the investigators permission to approach players as they were leaving the last green following a round of competition golf. The general objectives of the study were explained to players, and they were asked to take a questionnaire, which was completed in privacy and then returned to the investigators by reply-

76

Thomas and Over

paid post. Respondents were told that they could either complete the questionnaire anonymously or identify themselves by name or club membership number. Only 4 players refused to accept the questionnaire. Of the 400 questionnaires that were distributed, 172 were completed and returned (a response rate of 43%). To control for possible confounding developmental influences, 7 respondents aged less than 20 years were eliminated from the sample. Ages thus ranged from 20 years to 74 years (M = 48.15, SD = 12.03). Golf handicaps, as determined by the Australian Golf Union and used to adjust each player's score in every competition event, ranged from 5 to 27 (M = 15.67, SD = 5.30).

Procedure The substantive section of the questionnaire (the Golf Performance Survey) comprised 95 items assessing psychological and psychomotor skills, as well as level of involvement in golf. Most of these items were original, although some were adapted from sources such as the PPI (Loehr, 1982) and the PSIS (Mahoney et al., 1987), and others referenced component skills as described in golf instructional manuals. An initial version of the questionnaire was completed by 10 university staff who were active members of golf clubs. Feedback from pilot testing led to revision of some items and the instructions. The content covered by the 95 items forming the final version of the scale included a player's concentration, use of imagery, mental preparation, strategies and tactics, self-control, emotions and cognitions, psychomotor skills, commitment, and competitiveness. These attributes were assessed with direct reference to golf; thus, a respondent was asked to rate his psychological skills and competitiveness as a golfer rather than in general ("I am a mentally tough competitor in golf," "I am competitive when playing golf"). The respondent rated each item on a 5-point scale ranging from strongly disagree to strongly agree. In addition to completing the Golf Performance Survey, respondents provided biographical information, such as their age, height, weight, years of playing golf, the number of competition and practice rounds they currently played per month, and their current handicap. They also identified the number of shots they would take on average on each hole at the Pacific Golf Club, their average length of drive, and the percentage of times they would sink a putt from different distances. There also were questions assessing motives for playing golf. Forty of the 165 men who completed the Golf Performance Survey when it was initially distributed were randomly chosen to rate the 95 items assessing psychological and psychomotor skills in golf on a second occasion, 3 months after their initial ratings. Measures obtained from 36 of these players at the two time points were used to establish the test-retest reliability of the golf performance subscales.

Results Five items with markedly skewed distributions were excluded from data analyses. The remaining 90 items were then divided into three groupings (psychological and tactical skills, psychomotor skills, involvement in golf) on the basis of content. Dimensionality within each grouping was assessed through factor analysis based on ratings that the respondents had provided to each item. Principal axis factor analyses were conducted to explore the structure in each of the three data sets. To allow for intercorrelations between factors, oblique rotations were performed

Psychomotor Skills

77

using oblimin procedures. Alternative solutions were assessed on the basis of the discontinuity (scree) principle for eigenvalues, as well as in terms of the interpretability of rotated factors. In addition to the factor analyses, the process of subscale development involved scale reliability analyses to identify the itemtotal correlations and the internal consistency of each of the subscales. Data analysis thus yielded three scales, each with several subscales. These different scales will now be described, and attention then given to differences on the scales between lower handicap and higher handicap golfers.

Psychological Skills and Tactics in Golf Factor analysis of ratings to the items concerned with psychological skills and tactics in golf yielded five factors, accounting for 43% of variance. The items with factor loadings of .35 and item-total correlations of .30 or greater are listed in Table 1. Mean ratings on the 5-point scale, together with standard deviations, are shown for each of these items, as well as for the factor as a whole. Table 1 also reports the extent to which ratings of items were correlated with the age of respondents and their golf handicap. Factor 1 (10 items, alpha coefficient of .81) was labeled Negative Emotions and Cognitions; Factor 2 (9 items, alpha of .78), Mental Preparation; Factor 3 (4 items, alpha of .67), Conservative Approach; Factor 4 (3 items, alpha of .74), Concentration; and Factor 5 (4 items, alpha of .69), Striving for Maximum Distance. Test-retest reliabilities were .90 for Negative Emotions and Cognitions, .90 for Mental Preparation, .74 for Conservative Approach, .78 for Concentration, and .72 for Striving for Maximum Distance. The correlation coefficients reported in Table 1 indicate that as age increases, so does the tendency to play more conservatively, to strive for accuracy rather than to try to outdrive others, and to have fewer negative emotions and cognitions in playing golf. There was no evidence of an association between age and mental preparation or concentration. As levels of golf performance increase, there are likely to be corresponding increases in mental preparation and concentration during a game, the adoption of a less conservative approach, and fewer negative emotions and cognitions. No si@icant association was found between handicap and striving for maximum distance.

Psychomotor Skills in Golf Factor analysis of ratings to the items concerned with psychomotor skills yielded four factors accounting for 35% of variance, although the fourth factor is not considered further because of its low internal consistency (alpha coefficient of .45). The items with factor loadings of .35 and item-total correlations of .30 or greater are listed in Table 2. Mean ratings on the 5-point scale, together with standard deviations, are shown for each of these items, as well as for the factor as a whole. Table 2 also reports the extent to which ratings of each item were correlated with the age of respondents and their golf handicap. The first factor (13 items, alpha coefficient of .80) was labeled Automaticity; the second factor (3 items, alpha of .79), Putting Skill; and the third factor (8 items, alpha of .67), Seeking Improvement. Test-retest reliabilities were .92 for Automaticity, .85 for Putting Skill, and .78 for Seeking Improvement. The correlation coefficients reported in Table 2 indicate that Automaticity and Putting Skill are not significantly associated with age. However, the extent

78

Thomas and Over

Table 1 Item Loadings, Item-Total Subscale Correlations, Mean Ratings, and Correlations With Handicap and Age for Psychological Skills and Tactics Factor Item-total loading correlation M

Correlation with SD Handicap Age

Factor 1: Negative emotions and cognitions Choke under pressure in competitions Nervous when playing competitively Negative self-talk during competition Emotionally flat when playing poorly Anxious before hitting off first tee Thinking of past mistakes Angry and frustrated by a poor shot Unsettled by what others say or do Experience the yipps in putting Play better at practice than competition Negative emotions and cognitions subscale Factor 2: Mental preparation Work out where to place a shot Mentally rehearse each shot Consistent preshot routine Visualize where the ball will finish Mentally tough competitor Thought how best to play each hole Play best under pressure Visualize putting stroke and ball in hole Confident of playing to handicap Mental preparation subscale Factor 3: Conservative approach Lay up if unsure of clearing a hazard Prefer safe shots rather than take risks Play for the heart of a green Stick to what I can do, nothing new Conservative approach subscale Factor 4: Concentration Not easily distracted playing a shot Concentration is easily broken Unsettled by what others say or do Concentration subscale Factor 5: Striving for maximum distance Try to outdrive my opponent Hit the ball as far as I can when driving Satisfaction from longest drive in group Sacrifice distance for accuracy in driving Striving for maximum distance subscale

Note. Responses to positively loaded items were reflected prior to calculation of Concentration subscale scores. *p < .05. **p < .01. ***p < .001.

Psychomotor Skills

79

Table 2 Item Loadings, Item-Total Subscale Correlations, Mean Ratings, and Correlations With Handicap and Age for Psychomotor Skills Factor Item-total loading correlation

M

Correlation with SD Handicap Age

Factor 1: Automaticity Golf swing is well grooved Inconsistent driving from the tee Actions seem automatic playing a shot Lose more than one ball a round Swing so automatic, drive blindfolded Experimenting with alignment, position Some shots are too ambitious Drives depend on luck and chance Select wrong club for a shot Orthodox golf swing Underestimate distance to target Swing instinctively, little conscious effort Consistent in distances hit irons Automaticity subscale Factor 2: Putting skill Putting strength Good at reading greens Inconsistent in putting Putting skill subscale Factor 3: Seeking improvement Adjustments to grip or swing Lesson with a golf professional Leam by watching professionals play Practice putting before a round Practice round before major competition Watched golf instructional video Hit practice balls before a round Changed clubs to improve performance Seeking improvement subscale

Note. Responses to positively loaded items were reflectedprior to calculation of Putting Skill subscale scores. * p < .05. **p < .01. ***p c ,001.

to which players seek to improve their psychomotor skills does decline significantly with age. Automaticity is strongly related to golf performance. As Automaticity increases so, too, does the overall level of performance, and the golf handicap reduces. Scores on the Putting Skill and Seeking Improvement factors, however, were not significantly correlated with handicap.

Involvement in Golf Factor analysis of ratings to the items concerned with involvement in golf yielded two factors, but the second factor is not considered further because of its low

80

Thomas and Over

Table 3 Item Loadings, Item-Total Subscale Correlations, Mean Ratings, and Correlations With Handicap and Age for Involvement

Factor Item-total loading correlation M Factor 1: Commitment Playing golf well is important in life 0.64 Competitive when playing golf 0.52 Think about golf when not playing 0.47 Always trying to reduce handicap 0.47 Prefer social to competition golf -0.42 Commitment subscale

.56 .42 .43 .38 .34

3.54 3.82 3.82 3.79 2.30 3.73

Correlation with SD Handicap Age

0.99 0.81 0.94 0.95 1.00 0.62

-.16* -.29*** -.07 -.03 .24** -.24**

-.05 -.I1 -.I0 -.09 .09 -.I3

internal consistency (alpha coefficient of .58). The single remaining factor accounted for 16% of variance. The items with factor loadings of .35 and itemtotal correlations of .30 or greater are listed in Table 3. Mean ratings on the 5point scale, together with standard deviations, are shown for each item, as well as for the factor as a whole. Table 3 also reports the extent to which ratings of each item were correlated with the age of respondents and their golf handicap. This factor (five items, alpha coefficient of .67) was labeled Commitment. The test-retest reliability for the Commitment subscale was 35. The correlation coefficients reported in Table 3 indicate that the level of commitment to golf is significantly related to the players' proficiency level but not their age. As commitment increases, so too does the level of performance.

Correlations Between Psychological, Psychomotou, and Involvement Subscales Table 4 reports correlations between scores on the nine subscales developed following factor and reliability analyses. The number of statistically significant correlations contained in this table provides support for the decision to rotate factors obliquely. The general pattern of correlations indicates that the subscales do assess different attributes. The only intercorrelations above -40 were between Negative Emotions and Concentration (-.49), Mental Preparation and Automaticity (.43), and Mental Preparation and Commitment to golf (.40). These relationships make a good deal of sense and do not pose a threat to the construct validity of the subscales. As noted earlier, all subscales have acceptable internal consistency (alpha coefficients ranged from .67 to 3 1 , with a median value of .74). The values shown in parentheses in Table 4 indicate the test-retest reliability for each subscale. Because test-retest reliabilities ranged from .72 to .92, with a median value of .85, the measures can be considered highly stable at least over a period of 3 months.

Psychomotor Skills

81

Table 4 Correlations Between Scores on the Nine Golf Performance Subscales

1. Negative emotions and cognitions 2. Mental preparation 3. Conservative approach 4. Concentration 5. Striving for maximum distance 6. Automaticity 7. Putting skill 8. Seeking improvement 9. Commitment

Note. Test-retest reliability coefficients are shown in parentheses.

Differences Between Lower and Higher Handicap Golfers Each player in the sample had a handicap, which ranged from 5 to 27, reflecting the person's scores in competition over an extended period of time. The contrasts that follow are between lower handicap golfers (the more skilled players within the sample) and higher handicap golfers (the less skilled players in the sample). The aim in comparing these groups is to establish the extent to which scores on the psychological and psychomotor measures identified through factor analysis can be used to differentiate players performing at different levels in golf. Quintiles were used to obtain sufficient data on these measures from players differing widely in skill levels. Thus, the 34 respondents with golf handicaps of 11 or less (M = 8.62, SD = 1.92) were considered lower handicap golfers, and the 34 respondents with golf handicaps of 20 or greater (M = 23.29, SD = 2.63) were considered higher handicap golfers. Table 5 reports mean scores and standard deviations for these 2 groups on the subscales established through factor analysis. Comparison of the two groups across the nine measures through multivariate analysis of variance yielded a significant difference, F(9, 58) = 7.39, p < .001. Univariate tests revealed significantly higher mean scores for the more skilled golfers (the lower handicap group) than for the less skilled golfers (the higher handicap group) on the following measures: Mental Preparation, F(1,66) = 6.51, p < .05; Concentration, F(l, 66) = 6.08, p < .05; Automaticity, F(1,66) = 49.88, p < ,001; and Commitment, F(1, 66) = 5.07, p < .05. The lower handicap golfers were less prone than the higher handicap golfers to negative emotions and cognitions, F(1,66) = 8 . 5 8 , < ~ .01. The two groups did not differ significantly in mean scores on Conservative Approach, F(1,66) = 3.33, p > .05; Striving for Maximum Distance, F(1, 66) = 0.09, p > -05; Seeking Improvement, F(1, 66) = 3.07, p > .05; or Putting Skill, F(1, 66) = 0.00, p > ,05. Attention can be given not only to whether the difference between means on a measure is statistically significant but to effect size, the standardized mean difference (see Glass, McGaw, & Smith, 1981). The effect size values reported in Table 5 express the magnitude of the difference between groups on each measure as a standard score. Effect sizes were 0.60 or greater for four measures

82

Thomas and Over

Table 5 Mean Scores and Standard Deviations on the Nine Factors for the Two Groups, Together with Effect Size Values Lower handicap Factor

M

SD

Higher handicap M

SD

Effect size

Negative emotions and cognitions Mental preparation Conservative approach Concentration Striving for maximum distance Automaticity Putting skill Seeking improvement Commitment

(in descending order of magnitude of difference): Automaticity, Negative Emotions and Cognitions, Mental Preparation and Concentration. The magnitude of the difference on a measure between the two groups can also be referenced by the percentage of respondents in one group who have higher scores than the median respondent in the other group (see Rosenthal & Rubin, 1982). Assuming normal distributions, 84% of respondents in one group have higher scores than 50% of respondents in the other group on the measure in question when the effect size is 1.00. Analysis of the present data in these terms indicated that 96% of the lower handicap golfers had Automaticity scores above the median Automaticity score of the higher handicap golfers. The respective values were 73% for Mental Preparation and 73% for Concentration. Further, only 24% of golfers in the lower handicap group had higher scores for Negative Emotions and Cognitions than the median person in the higher handicap group. Multiple discriminant analysis was undertaken to determine the weighted combination of scores on the nine subscales that best differentiated the lower handicap players and the higher handicap players. The further interest was to establish accuracy in the classification of individual cases on the basis of the discriminant function. Multiple discriminant analysis was conducted with the nine subscales as the predictor measures and group membership as the criterion variable. The standardized canonical coefficients for the discriminant function were 0.98 for Automaticity, -0.43 for Mental Preparation, 0.35 for Seeking Improvement, -0.33 for Conservative Approach, 0.32 for Commitment, -0.29 for Negative Emotions and Cognitions, 0.07 for Striving for Maximum Distance, 0.06 for Concentration, and 0.03 for Putting Skill. This function permitted 56 of the 68 golfers (82%) to be classified correctly as a lower handicap player or as a higher handicap player on the basis of their responses to items within the nine subscales. Twenty-eight (82%) of the 34 lower handicap players were correctly classified, as were 28 (82%) of the 34 higher handicap players. In fact, only 2 of the 21 golfers with a handicap of 9 or less were incorrectly classified as a

Psychomotor Skills

83

higher handicap player, and only 3 of the 24 golfers with a handicap of 22 or more were incorrectly classified as lower handicap players.

Discussion The Golf Performance Survey developed in this study comprises nine subscales, including five on which mean scores differed significantly between lower handicap golfers and higher handicap golfers. The lower handicap (more skilled) golfers demonstrated greater mental preparation, better concentration, a higher level of automaticity, greater commitment, and fewer negative emotions and cognitions. These findings with amateur golfers are not unexpected in terms of research undertaken with professional golfers (Cohn, 1991; McCaffrey & Orlick, 1989). Successful amateurs, as well as professionals, demonstrate substantial commitment and mental preparation. In addition to giving greater thought to how best to play the course, skilled golfers report having a more consistent preshot routine that includes planning and mental rehearsal, particularly visualization. They regard themselves as mentally tough and play best under pressure. Not surprisingly, the low markers in the present study reported adopting a more competitive approach to golf and were keen to reduce their handicap even further. As well as being more able to focus attention when playing, the skilled golfers reported a greater capacity to regain focus following distraction. Consistent with such control of attentional focus, the lower handicap golfers were less liable than the higher handicap golfers to negative thoughts and feelings. The present data emphasize the importance of automaticity in skilled performance at golf. The low markers reported a well-grooved swing in which actions seem automatic rather than consciously controlled. In contrast, higher handicap golfers claimed that they continually make adjustments to their alignment, position, grip, or swing. They were more prone to errors of judgment in their shot making and generally reported a lower level of psychomotor control. Automaticity was associated with mental preparation and concentration,and the lack of negative thoughts and feelings. This pattern of relationships is consistent with the notion that once skills have been learned correctly, individuals need to trust in their own psychomotor abilities (Moore & Stevenson, 1991). Clearly, much of the differentiation between highly skilled and less skilled golfers was attributable to the levels of automaticity they reported. Although the Golf Performance Survey is a self-report instrument, behavioral measures can be used to validate scores on at least some of the subscales. For example, Automaticity, as measured by responses to items such as, "I lose no more than one ball per round," "I am consistent driving from the tee," and "My swing is so automatic that I can drive blindfolded" can be compared with performance on the golf course. Responses to other items in the Golf Performance Survey, including those tapping covert processes, such as visualization, selftalk, and emotional control, are more difficult to confirm. Additional research is needed to validate the subscale scores for these processes. Context is a further issue in validation of self-report. The golfers in the present study completed the Golf Performance Survey without direct reference to specific situations. It would be interesting to compare ratings under these conditions with ratings specific to context (e.g., practice vs. competition). A further possibility is to

84

Thomas and Over

obtain measures during play (e.g., establishing a player's response to an item such as "I get angry and frustrated when I play a poor shot" immediately following a specific shot). The Golf Performance Survey can be used in designing and evaluating training programs. As an initial step, a golfer's scores on each subscale would be compared with norms in order to identify the extent to which there are deficiencies in specific component skills. Training would focus on problem areas identified through such assessment, with the outcomes of training being evaluated on the basis not only of scores on the Golf Performance Survey but of performance on the golf course. The sport psychology literature points to interventions that might be included in training programs. For example, negative emotions and cognitions reflect arousal levels before and during competition, particularly after playing a poor shot, as well as negative self-talk. Several techniques can be used to regulate arousal (Harris & Williams, 1993) and control self-talk (Bunker, Williams, & Zinsser, 1993). Visualization during training in imagery and concentration must be tailored to the specific situations in which the skill will be performed (Lee, 1990; Syer & Connolly, 1987). There are guidelines available for imagery training programs (Vealey, 1990; Vealey & Walter, 1993), as well as techniques for improving concentration (Schmid & Peper, 1993). Once they are learned, psychological skills such as imagery and concentration need to be integrated in a preshot routine that is then developed through practice and used consistently in competition. Cohn (1990) has reviewed the efficacy of preperformance routines, and Boutcher (1990) provided a specific example of how psychological, psychophysiological, and behavioral components can be combined in a preshot routine for golf. A question of interest in further use of the Golf Performance Survey is whether differences in golf performance between men and women reflect differences in psychological and psychomotor component skills rather than simply in physical capacity (e.g., in strength). The manner in which performance in golf changes with age can also be studied through componential analysis, involving either longitudinal or cross-sectional comparisons. Studies of other aspects of psychomotor functioninghave suggested that molar (i.e., overall or global) performance can remain stable with age despite changes in component skills. In the case of typing, for example, deterioration in some component skills with age is compensated by improvement in other components (Salthouse, 1984). It would be interesting to determine for golf not only the extent to which molar performance (shots taken per round) changes with aging, but the manner in which changes occurring in overall prowess are mediated by deterioration or improvement in specific psychological and psychomotor skills of the type measured by the Golf Performance Survey.

References Albrecht, R.R., & FeItz, D.L. (1987). Generality and specificity of attention related to competitive anxiety and sport performance. Journal of Sport Psychology, 9,231-248. Bandura, A. (1986). Social foundations of thought and action. Englewood Cliffs, NJ: Prentice Hall. Boutcher, S.H. (1990). The role of performance routines in sport. In J.G. Jones & L. Hardy (Eds.), Shess andpe$orrnance in sport (pp. 231-245). Chichester, England: Wiley.

Psychomotor Skills

85

Bunker, L., Williams, J.M., & Zinsser, N. (1993). Cognitive techniques for improving performance and building confidence. In J.M. Williams (Ed.), Applied sportpsychology (2nd ed., pp. 225-242). Palo Alto, CA: Mayfield. Cohn, P.J. (1990). Preperformance routines in sport: Theoretical support and practical applications. The Sport Psychologist, 4, 301-312. Cohn, P.J. (1991). An exploratory study on peak performance in golf. The Sport Psychologist, 5, 1-14. Cohn, P.J., Rotella, R.J., & Lloyd, J.W. (1990). Effects of acognitive-behavioral intervention on the preshot routine and performance in golf. 7'he Sport Psychologist, 4, 3347. Fisher, A.C. (1984). New directions in sport personality research. In J.M. Silva & R.S. Weinberg (Eds.), Psychological foundations of sport (pp. 70-80). Champaign, IL,: Human Kinetics. Glass, G.V., McGaw, B., & Smith, M.L. (1981). Meta-analysis in social research. Beverly Hills, CA: Sage. Greenspan, M., Murphy, S., Tammen, V.,& Jowdy, D. (1989). Eflects of athlete achievement level and test administration on the Psychological Skills Inventory for Sports (PSIS).Paper presented at the annual conference of the Association for the Advancement of Applied Sport Psychology, Nashua, NH. Harris, D.V., & Williams, J.M. (1993). Relaxation and energizing techniques for regulation of arousal. In J.M. Williams (Ed.), Applied sport psychology (2nd ed., pp. 185199). Palo Alto, CA: Mayfield. Lee, C. (1990). Psyching up for a muscular endurance task: Effects of image content on performance and mood state. Journal of Sport & Exercise Psychology, 12, 66-73. Lesser, M., & Murphy, S. (1988). The Psychological Skills Inventory for Sport (PSIS): Normative and reliability data. Paper presented at the annual conference of the American Psychological Association, Atlanta, GA. Loehr, J.E. (1982). Mental toughness training for sports. Lexington, MS: Stephen Greene Press. Mahoney, M.J., Gabriel, T.J., & Perkins, T.S. (1987). Psychological skills and exceptional athletic performance. The Sport Psychologist, 1, 181-199. McCaffrey, N., & Orlick, T. (1989). Mental factors related to excellence among top professional golfers. International Journal of Sport Psychology, 20, 256-278. Mischel, W. (1977). The interaction of person and situation. In D. Magnusson & N.S. Endler (Eds.), Personality at the crossroads: Current issues in interactionalpsychology (pp. 333-352). Hillsdale, NJ: Erlbaum. Moore, W.E., &Stevenson, J.R. (1991). Understanding trust in the performance of complex automatic sport skills. The Sport Psychologist, 5, 281-289. Nideffer, R.M. (1976). Test of Attentional and Interpersonal Style. Journal of Personality and Social Psychology, 34, 394-404. Nideffer, R.M. (1985).Athletes' guide to mental training. Champaign, IL: Human Kinetics. Ogilvie, B. (1991, November). Critical use ofpsychological tests in sport. Keynote address, First Asian South Pacific Association of Sports Psychology International Congress, Melbourne, Australia. Rosenthal, R., & Rubin, D.B. (1982). A simple, general purpose display of magnitude of experimental effect. Journal of Educational Psychology, 74, 166-169. Salthouse, T.A. (1984). Effects of age and skill in typing. Journal of Experimental Psychology: General, 113, 345-371. Schmid, A., & Peper, E. (1993). Training strategies for concentration. In J.M. Williams (Ed.), Applied sport psychology (2nd ed., pp. 262-273). Palo Alto, CA: Mayfield.

86

Thomas and Over

Schulz, R., & Cumow, C. (1988). Peak performance and age among superathletes: Track and field, swimming,baseball, tennis, and golf. Journal of Gerontology:Psychological Sciences, 43, 113-120. Silva, J.M. (1984). Personality and sport performance: Controversy and challenge. In J.M. Silva & R.S. Weinberg (Eds.), Psychological foundations of sport (pp. 59-69). Champaign, IL: Human Kinetics. Singer, R.N. (1988). Psychological testing: What value to coaches and athletes? lnternational Journal of Sport Psychology, 19, 87-106. Summers, J.J., Miller, K., & Ford, S. (1991). Attentional style and basketball performance. Journal of Sport & Exercise Psychology, 13, 239-253. Syer, J., & Connolly, C. (1987). Sporting body sporting mind. Englewood Cliffs, NJ: Prentice Hall. Tammen, V.V., Murphy, S.M., & Jowdy, D. (1990). Reevaluating the Psychological Skills Inventory for Sports: Factor analysis and implications. Paper presented at the annual conference of the North American Society of Psychology of Sport and Physical Activity, Asilomar, CA. Van Schoyck, R.S., & Grasha, A.F. (1981). Attentional style variations and athletic ability: The advantage of a sport-specific test. Journal of Sport Psychology, 3, 149-165. Vealey, R.S. (1990). Inner coaching through mental imagery. Sport Psychology Training Bulletin, 2(2), 1-7. Vealey, R.S., & Walter, S.M. (1993). Imagery training for performance enhancement and personal development. In J.M. Williams (Ed.), Applied sport psychology (2nd ed., pp. 200-224). Palo Alto, CA: Mayfield.

Acknowledgments We are grateful to members of the Pacific Golf Club in Brisbane, Australia, who participated in the survey on which the data analysis is based. Manuscript submitted: September 14, 1992 Revision received: August 10, 1993