The use of physiological, anthropometric, and skill data to predict selection in a talent-identified junior volleyball squad

Transcription

1 Journal of Sports Sciences, October 2007; 25(12): The use of physiological, anthropometric, and skill data to predict selection in a talent-identified junior volleyball squad TIM GABBETT 1, BORIS GEORGIEFF 1, & NATHAN DOMROW 2 1 Athlete and Coach Support Services, Queensland Academy of Sport, Brisbane and 2 Office of Economic and Statistical Research, Queensland Treasury, Brisbane, QLD, Australia (Accepted 30 December 2006) Abstract The aim of this study was to determine whether physiological, anthropometric, and skill test results could discriminate between junior volleyball players of varying ability. Twenty-eight junior volleyball players competed for selection in a talentidentification volleyball programme. Participants underwent measurements of stature, standing reach stature, body mass, skinfold thickness, overhead medicine ball throw, vertical jump, spike jump, 5-m and 10-m speed, T test agility, maximal aerobic power, and passing, setting, serving, and spiking technique and accuracy. A discriminant analysis was conducted on the selected and non-selected groups to obtain a regression equation that could be used to predict selection in junior volleyball squads based on the dependent variables. Passing and serving technique were the only significant variables included in the discriminant analysis. Cross-validation results showed that 17 of 19 selected players (89.5%) and 5 of 9 nonselected players (55.6%) were correctly classified into selected and non-selected groups, respectively, providing an overall predictive accuracy of 78.6%. The results of this study demonstrate that selected skill test results (i.e. subjective coach evaluations of passing technique and serving technique), but not physiological and anthropometric data, discriminate between successful and unsuccessful talent-identified junior volleyball players. These results demonstrate the importance of developing passing and serving technique in talent-identified junior volleyball players. Keywords: Discriminant analysis, talent identification and development, physical fitness, Talent Search volleyball programme Introduction Volleyball is an intermittent sport that requires players to participate in frequent short bouts of high-intensity exercise, followed by periods of low-intensity activity (Kunstlinger, Ludwig, & Syegemann, 1987; Viitasalo et al., 1987). The highintensity bouts of exercise, coupled with the total duration of the match (*90 min), requires players to have well-developed aerobic and anaerobic alactic (ATP-CP) energy systems (Polglaze & Dawson, 1992; Viitasalo et al., 1987). Considerable demands are also placed on the neuromuscular system during the various sprints, jumps (blocking and spiking), and high-intensity court movement that occurs repeatedly during competition (Hosler, Morrow, & Jackson, 1978). As a result, volleyball players require well-developed speed, agility, upper-body and lowerbody muscular power, and maximal aerobic power ( _V O 2max ). Several studies have documented the physiological and anthropometric characteristics of volleyball players (Fleck, Case, Puhl, & Van Handle, 1985; Hakkinen, 1993; Hosler et al., 1978; Spence, Disch, Fred, & Coleman, 1980), with the fitness of players typically increasing as the playing standard is raised (Gabbett & Georgieff, 2007; Smith, Roberts, & Watson, 1992; Thissen-Milder & Mayhew, 1991). Smith et al. (1992) compared physical, physiological, and performance characteristics of national- and college-standard volleyball players and found significantly higher block and spike jumps, 20-m speed, and _V O 2max in the national-standard players, suggesting that physiological capacities play an important role in the preparation and selection of elite volleyball players (Smith et al., 1992). In addition, Thissen-Milder and Mayhew (1991) demonstrated that selected physiological and anthropometric characteristics could successfully discriminate among freshman, junior varsity, and varsity volleyball teams Correspondence: T. Gabbett, Athlete and Coach Support Services, PO Box 956, Sunnybank, Brisbane, QLD 4109, Australia. tim.gabbett@srq.qld.gov.au ISSN print/issn X online Ó 2007 Taylor & Francis DOI: /

2 1338 T. Gabbett et al. and starting and non-starting players. Changes in the physiological and anthropometric characteristics of volleyball players in response to training and over the course of a season have also been documented (Fardy, Hritz, & Hellerstein, 1976; Franks & Moore, 1969; Gabbett & Georgieff, 2005; Hascelik, Basgoze, Turker, Narman, & Ozker, 1989). Studies of the effect of volleyball and physical conditioning training on the physiological and anthropometric characteristics of players are equivocal, with reports of increased (Fardy et al., 1976; Franks & Moore, 1969; Hascelik et al., 1989), decreased (Hakkinen, 1993) or unchanged fitness (Gabbett & Georgieff, 2005) in response to training. Improvements in _V O 2max (Franks & Moore, 1969; Hascelik et al., 1989), speed (Franks & Moore, 1969), strength (Hascelik et al., 1989), and visual reaction time (Hascelik et al., 1989) have been reported following 5 10 weeks of volleyball and physical conditioning training. However, recent evidence has demonstrated unchanged speed, agility, lower-body muscular power, and _V O 2max over a season in junior elite volleyball players (Gabbett & Georgieff, 2005). In addition, it was recently shown that volleyball training significantly improved the skill levels of talent-identified volleyball players, without significantly altering skinfold thickness, upper-body muscular power, lowerbody muscular power or _V O 2max (Gabbett et al., 2006). Collectively, these findings (Gabbett & Georgieff, 2005; Gabbett et al., 2006) suggest that skill execution is at least as important as physiological and anthropometric characteristics for predicting success in junior volleyball players. Although physiological and anthropometric characteristics have been used to discriminate between athletes of varying ability for other team (and fast ball) sports such as rugby union (Rigg & Reilly, 1988), rugby league (Gabbett, 2002a, 2002b; Gabbett & Herzig, 2004), soccer (Hoare & Warr, 2000), and Australian football (Keogh, 1999), no study has determined if the physiological and anthropometric characteristics of volleyball players influence selection into talent-identification volleyball squads. In addition, the discriminatory ability of players skill levels has not been investigated. With this in mind, the purpose of this study was to determine if physiological, anthropometric, and skill test results could discriminate between junior volleyball players of varying ability, competing for selection in a talent-identification volleyball programme. Methods Participants Twenty-eight junior volleyball players (mean age, 15.5 years, s ¼ 1.0) participated in this study. An initial open invitation to trial for the programme was extended to students from all secondary schools within the Brisbane metropolitan area. After the initial trial, athletes were selected into a preliminary training squad based on subjective coaching opinions of their movement coordination in game-specific tasks. Alhough physiological and anthropometric testing was performed at this trial, the data were not used in the selection process. Following selection into the preliminary training squad, all athletes participated in volleyball on a trial basis for 3 months. All players were competing for full selection within the Queensland Academy of Sport Talent Search volleyball programme. This programme identifies young athletes deemed to have the necessary physiological and anthropometric characteristics (e.g. stature, standing reach stature, muscular power, speed, agility, and maximal aerobic power) for volleyball success and places those athletes in a high-performance coaching environment, where they are provided with specialized volleyball coaching. Although the participants had limited volleyball experience, they had participated in a wide range of sports (e.g. swimming, track and field, martial arts, mountain biking, tennis, netball, basketball, hockey, touch football, and rugby union) before attempting selection for the Talent Search volleyball programme. Twenty-four athletes (85.7%) had participated in one sport, while four athletes (14.3%) had participated in two or more sports. All participants received a clear explanation of the study, including the risks and benefits of participation, and written parental or guardian consent was obtained. The Institutional Review Board for Human Investigation approved all experimental procedures. Skill testing battery Players underwent accuracy and technique assessments of spiking, passing, setting, and serving skills in an indoor stadium (Gabbett & Georgieff, 2006). After a standardized warm-up, players performed six trials of each skill. Two 37-mm digital cameras (Sony, DCR-TRV 950), positioned approximately 5 m from the player, were used to film each skill. For spiking and setting skills, players were filmed from the side and behind, whereas serving and passing skills were filmed from the side and front of the player. The players accuracy was based on their ability to hit specific targets. The players technique was subjectively evaluated from video footage by two expert coaches using standardized technical criteria (Table I). Neither of the coaches was involved in the selection process and each was blinded to the amount of volleyball-specific training that players had performed. Coaches used an 7-point (1 7) Likert scale to assess the player s technique.

3 Predicting selection in junior volleyball 1339 Spiking. A target (2 m high and 1 m wide) was placed upon a wall, 7 m from the player (see Figure 1). The target was divided into five 20-cm segments. Players were instructed to throw the ball approximately 1 2 m in the air and hit the ball towards the target. If players were able to hit the middle 20-cm segment, they were awarded 5 points. Three points were awarded for successfully hitting the two 20-cm segments on either side of the middle segment, while 1 point was awarded for hitting the two outer 20-cm segments of the target. A score of 2 Table I. Criteria used by coaches to assess technical skill of volleyball players. Skill Spiking Setting Serving Passing Criteria Identify height, speed and location of set Angle approach to maximize hitting zone Build speed and increase momentum of approach Contact ball high in front of hitting shoulder Fast arm swing that follows through past the contact point Feet, hips, and shoulders facing target Hips forward and an upright body position Ball set from forehead and above Follow through to target (arms and legs) A disciplined and consistent routine before each serve Controlled toss in front of the hitting shoulder Solid contact in the centre of the ball (without spinning) Low trajectory Feet slightly wider than shoulder width Knees and back slightly bent Arms away from the body Elbows locked and shoulders rotated forward Arms tilted towards the target and 4 points were awarded if players successfully hit the target between the 1 point and 3 point segments, and 3 point and 5 point segments, respectively. The aggregate from six trials was recorded as the players accuracy score. Passing. The passing ability of the players was evaluated by determining their ability to return a pass to a target positioned at the net, 2 m from the right-hand sideline. The target dimensions were 1.6 m long and 2.3 m wide (see Figure 2). This target was chosen as it was the approximate position the setter would transition to during a match. A coach, positioned in the service position, approximately 1 m above the ground and 10 m from the receiving player, threw an overhead pass to the receiving player. Players were required to pass (dig) the ball to another player standing with arms extended above their head, in the target area. Players who successfully passed the ball to the player in the target area were awarded 2 points. A second target area was created for passes that did not make the main target area, but would be likely to reach another player during a competitive match. The second target extended from the right-hand sideline and was 3 m long and 4.1 m wide. Players who successfully passed the ball within the second target area were awarded 1 point. Finally, a pass that did not reach either of the target areas was awarded no points. The aggregate from six trials was recorded as the players accuracy score. Setting. The setting ability of players was evaluated by determining their ability to set to a target positioned next to the net at net height, 5.5 m from the setting player (see Figure 3). This target was chosen as it was Figure 1. Schematic illustration of spiking task. Figure 2. Schematic illustration of passing task.

4 1340 T. Gabbett et al. Figure 3. Schematic illustration of setting task. the approximate position a spiking player would contact the ball following their approach and take-off during a match. A coach, positioned approximately 5 m from the setting player, threw an overhead pass to the setting player. Players were required to set the ball to a target, 80 cm in diameter. Players who successfully set the ball through the target were awarded 3 points. Balls that hit the outside edge of the target, but did not go through the target, were awarded 2 points. Players who set the ball within 2.3 m of the net (and therefore 1.5 m of the target) were awarded 1 point. Balls that did not reach the target areas were awarded no points. The aggregate from six trials was recorded as the players accuracy score. Serving. Serving accuracy was determined as the ability of players to serve in the court from a service position. Given that the players had limited exposure to volleyball before the study, it was considered unreasonable for players to possess the coordination to hit specific service targets (other than an entire court) (i.e. 9 m width). The aggregate from six trials was recorded as the players accuracy score. Fitness testing battery Standard anthropometry (stature, standing reach stature, body mass, and sum of seven skinfolds), overhead medicine ball throw, vertical jump, spike jump, speed (5-m and 10-m sprints), agility ( T test), and maximal aerobic power (multi-stage fitness test) were the physiological and anthropometric tests selected. Players were instructed to refrain from strenuous exercise for at least 48 h prior to the fitness testing session and consume their normal pretraining diet before the testing session. The testing session began with anthropometric measurements. Players then underwent measurements of overhead medicine ball throw, vertical jump, spike jump, speed (5-m and 10-m sprints), and agility ( T test) measurements. Participants performed three trials for the speed, agility, vertical jump, spike jump, and overhead medicine ball throw tests, with a recovery of approximately 3 min between trials. Players were encouraged to perform low-intensity activities and stretches between trials. Upon completion of the respective tests, the field-testing session concluded with players performing the multi-stage fitness test (estimated maximal aerobic power). All fitness testing was performed in an indoor stadium. Anthropometry. As an estimate of adiposity, skinfold thickness was measured at seven sites using a Harpenden skinfold caliper. Biceps, triceps, subscapular, supraspinale, abdomen, thigh, and calf on the right side were the seven sites selected. The exact positioning of each skinfold measurement was in accordance with procedures described by Norton et al. (2000). Stature was measured using a stadiometer, and body mass was measured using calibrated digital scales (A&D Company Ltd., Tokyo, Japan). Standing reach stature was measured using a Yardstick vertical jump device (Swift Performance Equipment, NSW, Australia). Players were requested to stand with feet flat on the ground, extend their arm and hand, and mark the standing reach stature. Overhead medicine ball throw. For the overhead medicine ball throw, players stood one step behind a line marked on the ground facing the throwing direction, with a 3-kg medicine ball held in both hands behind the head. Players were instructed to plant the front foot with the toe behind the line and throw the medicine ball overhead as far forward as possible. Each throw was measured from inside the line to the nearest mark made by the fall of the medicine ball. Throwing distance was measured to the nearest centimetre with the best value obtained from three trials used as the overhead throw score.

5 Predicting selection in junior volleyball 1341 Vertical jump. The vertical jump and spike jump tests were performed using a Yardstick vertical jump device (Swift Performance Equipment, NSW, Australia). Players were requested to stand with feet flat on the ground, extend their arm and hand, and mark the standing reach height. After assuming a crouch position, each participant was instructed to spring upward and touch the Yardstick device at the highest possible point. No specific instructions were given about the depth or speed of the countermovement. Vertical jump height was calculated as the distance from the highest point reached during standing and the highest point reached during the vertical jump. Vertical jump height was measured to the nearest centimetre with the best value obtained from three trials used as the vertical jump score. Spike jump. Similar procedures to the vertical jump were used for the spike jump. Players were requested to stand with feet flat on the ground, extend their arm and hand, and mark the standing reach height. Players were then instructed to take a run-up or spike approach and leap as high as possible off both legs, displacing as many vanes on the Yardstick as possible. Spike jump height was calculated as the distance from the highest point reached during standing and the highest point reached during the spike jump. Spike jump height was measured to the nearest centimetre with the best value obtained from three trials used as the spike jump score. Speed. The running speed of players was evaluated at 5 m and 10 m over a 10-m sprint effort using dualbeam electronic timing gates (Swift Performance Equipment, NSW, Australia). The timing gates were positioned 5 m and 10 m from a pre-determined starting point. Players were instructed to run as quickly as possible along the 10-m distance from a standing start. Speed was measured to the nearest 0.01 s with the best value obtained from three trials used as the speed score. Agility. The agility of players was evaluated using a T test (Hoffman, Maresh, Armstrong, & Kraemer, 1991) using dual-beam electronic timing gates (Swift Performance Equipment, NSW, Australia). Four cones were placed 5 m apart in the shape of an inverted T. Players were instructed to run as quickly as possible along the agility run. Agility times were measured to the nearest 0.01 s with the best value obtained from three trials used as the agility score. Maximal aerobic power. Maximal aerobic power was estimated using the multi-stage fitness test (Ramsbottom, Brewer, & Williams, 1988). Players were required to run back and forth (i.e. shuttle run) along a 20-m track, keeping in time with a series of signals on a compact disk. The frequency of the audible signals (and hence running speed) was progressively increased, until players reached volitional exhaustion. Maximal aerobic power ( _V O 2max ) was estimated using regression equations described by Ramsbottom et al. (1988). Statistical analysis Differences in the anthropometric characteristics, overhead medicine ball throw, vertical jump, spike jump, speed, agility, _V O 2max, and skill levels of the selected and non-selected players were compared using an independent t-test. A linear discriminant analysis was also conducted on the two groups to obtain a regression equation that could be used to predict selection in junior volleyball squads on the basis of the dependent variables. For this analysis, a homogenous variance and multivariate normal withingroup distribution was assumed (Lachenbruch, 1975). Using a pooled covariance matrix (which yields a linear function) and a proportional prior distribution, linear discriminant analysis was undertaken. Based on awilks lambdasignificancelevelof0.05,astepwise selection method with a 0.15 selection probability criterion was used (Everitt & Dunn, 1991). Statistical significance was set at P and all data are reported as means and standard deviations (s). Results The skill, physiological, and anthropometric characteristics of selected and non-selected players are shown in Table II. Selected players had greater (P ¼ 0.02 to 0.001) passing accuracy, and spiking, serving, and passing technique, than non-selected players. There were no differences (P ¼ 0.08 to 0.49) between selected and non-selected players for any of the physiological or anthropometric characteristics. Equations (1) and (2) show the discriminant analysis equation that were developed to predict which players would be successful and unsuccessful, respectively, in being selected for the volleyball squad. Subjective coach evaluations of passing technique and serving technique were the only significant (P ¼ to 0.002) variables included in the discriminant analysis. Discriminant analysis to predict selection in junior volleyball: Discriminant score ¼ 15:64 þ 3:93 passing technique þ 4:01 serving technique ð1þ

6 1342 T. Gabbett et al. Table II. Physiological, anthropometric, and skill results for selected and non-selected volleyball players (mean + s) Selected (n ¼ 19) Discriminant analysis to predict non-selection in junior volleyball: Discriminant score ¼ 8:83 þ 2:56 passing technique þ 3:08 serving technique Non-selected (n ¼ 9) Accuracy Spiking accuracy Serving accuracy Setting accuracy Passing accuracy * Technique Spiking technique * Serving technique * Setting technique Passing technique * Anthropometric characteristics Body mass (kg) Stature (m) Standing reach stature (cm) Sum of skinfolds (mm) Physiological characteristics Vertical jump (cm) Spike jump (cm) Overhead-medicine ball throw (m) 5-m sprint (s) m sprint (s) Agility (s) _V O 2max (ml kg 71 min 71 ) *Significantly different (P ) from non-selected athletes. ð2þ Cross-validation results were calculated to identify misclassified observations. Equation (1) and (2) correctly predicted 17 of 19 selected players and 5 of 9 non-selected players, respectively, 22 of 28 players in total. The discriminant analysis corresponded to an overall predictive accuracy of 78.6% for all players, and an accuracy of 89.5% and 55.6% for the selected and non-selected players, respectively. Discussion The present study is the first to investigate if physiological, anthropometric, and skill test results discriminate between junior volleyball players of varying playing ability. The results of this study demonstrate that selected skill test results (i.e. subjective coach evaluations of passing technique and serving technique), but not physiological and anthropometric data, discriminate between successful (i.e. selected) and unsuccessful (i.e. non-selected) talent-identified junior volleyball players. These results demonstrate the importance of developing passing and serving technique in talent-identified junior volleyball players. The discriminant analysis correctly predicted 17 of 19 selected players and 5 of 9 non-selected players, that is 22 of 28 players in total. The prediction equations corresponded to an overall accuracy of 78.6% for all players, and an accuracy of 89.5% and 55.6% for the selected and non-selected players, respectively. These findings demonstrate that by using the classification functions, it is possible to discriminate with better than chance accuracy whether junior volleyball players will be correctly selected into a talent-identification squad using assessments of their passing and serving technique. However, while the classification functions correctly selected 17 of 19 players (i.e. 89.5%), the accuracy of the classification functions for predicting nonselection was lower (i.e. 5 of 9 players, 55.6%). Given that the discriminant analysis prediction equations were developed from one selection trial, it is likely that repeated selection trials would lead to more robust conclusions and a higher predictive accuracy for non-selected players. In addition, these findings demonstrate that while above average passing and serving technique may facilitate selection, factors in addition to, or other than, passing and serving technique may influence non-selection. Significant differences were detected between selected and non-selected junior volleyball players for spiking, serving, and passing technique and passing accuracy. In addition, passing technique and serving technique were the only variables that contributed significantly to the discriminant analysis. These findings are in partial agreement with those of others (Thissen-Miler & Mayhew, 1991) who reported differences among freshmen, junior varsity, and varsity-standard volleyball players for passing and spiking skills. However, the finding of similar physiological and anthropometric characteristics between groups is in conflict with other studies that have found a progressive improvement in physical qualities as the playing standard increases (Gabbett & Georgieff, 2007; Smith et al., 1992; Thissen- Milder & Mayhew, 1991). Indeed, in a recent study Gabbett and Georgieff (2007) reported significant differences among junior national, state, and novice volleyball players for stature, standing reach stature, skinfold thickness, vertical jump, spike jump, agility, and estimated maximal aerobic power. The discrepant findings between the present and previous studies is difficult to reconcile, but could reflect differences in coaching priorities between playing populations with elite players requiring higher fitness and talent-identified players requiring a skill

7 Predicting selection in junior volleyball 1343 coaching emphasis to enhance skill acquisition and development. Although the physiological and anthropometric characteristics of players were similar between groups, and none of the physiological and anthropometric parameters contributed significantly to the discriminant analysis, this does not suggest that physical fitness is unimportant to the competitive performances of junior volleyball players. Indeed, while passing technique and serving technique were the only variables that contributed significantly to the discriminant analysis model, the prediction equations are developed from the interaction of all physiological, anthropometric, and skill variables. These findings suggest that physiological and anthropometric monitoring should be included in any testing of junior volleyball players to provide coaches with objective feedback on the individual strengths and weaknesses of players. Perhaps more importantly, the ability to perform skillfully is constrained by physiological limitations. For example, a player may have good passing and serving technique, but if unable to jump high to hit the ball over the net (i.e. demonstrate good muscular power), good technique will not necessarily lead to improved performance. Therefore, players require a combination of well-developed physiological (e.g. muscular power and agility) and anthropometric (e.g. stature) characteristics to effectively perform the skills of volleyball. While the present results clearly demonstrate that passing and serving technique are important to predicting selection in a talent-identified junior volleyball squad, the prediction equations developed in this study may not be directly applicable to elite volleyball players, as any developed model is best suited to the population from which it is derived (Keogh, 1999). Given the popularity of volleyball in European countries, it is possible that the physiological and anthropometric characteristics of players are more highly developed, and therefore play a more influential role in player selection. In addition, while the statistical analysis determined the most important factors for selecting individual players into the talentidentification squad, these factors may not be the most important factors for choosing the most competitive team. Unmeasured factors, such as ability to cooperate in a team environment or leadership capabilities, could be equally important. In conclusion, the present study investigated if physiological, anthropometric, and skill test results could discriminate between junior volleyball players of varying ability, competing for selection in a talentidentification volleyball programme. The results demonstrate that selected skill test results (i.e. subjective coach evaluations of passing technique and serving technique), but not physiological and anthropometric data, discriminate between successful (i.e. selected) and unsuccessful (i.e. non-selected) talent-identified junior volleyball players. These results demonstrate the importance of developing passing and serving technique in talent-identified junior volleyball players. References Everitt, B. S., & Dunn, G. (1991). Applied multivariate data analysis (pp ). London: Edward Arnold. Fardy, P. S., Hritz, M. G., & Hellerstein, H. K. (1976). Cardiac responses during women s intercollegiate volleyball and physical fitness changes from a season of competition. Journal of Sports Medicine, 16, Fleck, S. J., Case, S., Puhl, J., & Van Handle, P. (1985). Physical and physiological characteristics of elite women volleyball players. Canadian Journal of Applied Sports Sciences, 10, Franks, B. D., & Moore, G. C. (1969). Effects of calisthenics and volleyball on the AAHPER fitness test and volleyball skill. Research Quarterly, 40, Gabbett, T. J. (2002a). Influence of physiological characteristics on selection in a semi-professional rugby league team: A case study. Journal of Sports Sciences, 20, Gabbett, T. J. (2002b). Physiological characteristics of junior and senior rugby league players. British Journal of Sports Medicine, 36, Gabbett, T., & Georgieff, B. (2005). Physiological characteristics of elite junior volleyball players over a competitive season. Strength and Conditioning Coach, 13, 2 7. Gabbett, T. J., & Georgieff, B. (2006). The development of a standardized skill assessment for junior volleyball players. International Journal of Sports Physiology and Performance, 1, Gabbett, T., & Georgieff, B. (2007). Physiological and anthropometric characteristics of junior national, state, and novice volleyball players. Journal of Strength and Conditioning Research. Gabbett, T., Georgieff, B., Anderson, S., Cotton, B., Savovic, D., & Nicholson, L. (2006). Changes in skill and physical fitness following training in talent-identified volleyball players. Journal of Strength and Conditioning Research, 20, Gabbett, T. J., & Herzig, P. J. (2004). Physiological characteristics of junior elite and sub-elite rugby league players. Strength and Conditioning Coach, 12, Hakkinen, K. (1993). Changes in physical fitness profile in female volleyball players during the competitive season. Journal of Sports Medicine and Physical Fitness, 33, Hascelik, Z., Basgoze, O., Turker, K., Narman, S., & Ozker, R. (1989). The effects of physical training on physical fitness tests and auditory and visual reaction times of volleyball players. Journal of Sports Medicine and Physical Fitness, 29, Hoare, D. G., & Warr, C. R. (2000). Talent identification and women s soccer: An Australian experience. Journal of Sports Sciences, 18, Hoffman, J. R., Maresh, C. M., Armstrong, L. E., & Kraemer, W. J. (1991). Effects of off-season and in-season resistance training programs on a collegiate male basketball team. Journal of Human Muscle Performance, 1, Hosler, W. W., Morrow, J. R., & Jackson, A. S. (1978). Strength, anthropometric, and speed characteristics of college women volleyball players. Research Quarterly, 49, Keogh, J. (1999). The use of physical fitness scores and anthropometric data to predict selection in an elite under 18 Australian rules football team. Journal of Science and Medicine in Sport, 2,

8 1344 T. Gabbett et al. Kunstlinger, U., Ludwig, H. G., & Stegemann, J. (1987). Metabolic changes during volleyball matches. International Journal of Sports Medicine, 8, Lachenbruch, P. A. (1975). Discriminant analysis. New York: Hafner. Norton, K., Marfell-Jones, M., Whittingham, N., Kerr, D., Carter, L., Saddington, K. et al. (2000). Anthropometric assessment protocols. In C. J. Gore (Ed.), Physiological tests for elite athletes (pp ). Champaign, IL: Human Kinetics. Polglaze, T., & Dawson, B. (1992). The physiological requirements of the positions in state league volleyball. Sports Coach, 15, Ramsbottom, R., Brewer, J., & Williams, C. (1988). A progressive shuttle run test to estimate maximal oxygen uptake. British Journal of Sports Medicine, 22, Rigg, P., & Reilly, T. (1988). A fitness profile and anthropometric analysis of first and second class rugby union players. In T. Reilly, A. Lees, K. Davids, & W. J. Murphy (Eds.), Science and football: Proceedings of the First World Congress of Science and Football (pp ). London: E & FN Spon. Smith, D. J., Roberts, D., & Watson, B. (1992). Physical, physiological and performance differences between Canadian national team and universiade volleyball players. Journal of Sports Sciences, 10, Spence, D. W., Disch, J. G., Fred, H. L., & Coleman, A. E. (1980). Descriptive profiles of highly skilled women volleyball players. Medicine and Science in Sports and Exercise, 12, Thissen-Milder, M., & Mayhew, J. L. (1991). Selection and classification of high school volleyball players from performance tests. Journal of Sports Medicine and Physical Fitness, 31, Viitasalo, J., Rusko, H., Pajala, O., Rahkila, P., Ahila, M., & Montonen, H. (1987). Endurance requirements in volleyball. Canadian Journal of Applied Sports Sciences, 12,

9

10