Maturity-associated variation in sport-specific skills of youth soccer players aged years

Journal of Sports Sciences, May 2005; 23(5): 515 522 Maturity-associated variation in sport-specific skills of youth soccer players aged 13 15 years ROBERT M. MALINA 1, SEAN P. CUMMING 2, ANTHONY P. KONTOS 3, JOEY C. EISENMANN 4, BASIL RIBEIRO 5, & JOÃO AROSO 5 1 Research Professor, Tarleton State University, Stephenville, TX, USA, 2 Department of Psychology, University of Washington, Seattle, WA, USA, 3 Human Performance Center, University of New Orleans, New Orleans, LA, USA, 4 Department of Health and Human Performance, Iowa State University, Ames, IA, USA, and 5 Instituto Superior da Maia, Castero da Maia, Avioso S. Pedro, Portugal (Accepted 7 May 2004) Abstract The aim of this study was to estimate the contribution of experience, body size and maturity status to variation in sportspecific skills of adolescent soccer players. The participants were 69 players aged 13.2 15.1 years from three clubs that competed in the highest division for their age group. Height and body mass were measured and stage of pubic hair development was assessed at clinical examination. Years of experience in football was obtained at interview. Six football skill tests were administered: ball control with the body, ball control with the head, dribbling with a pass, dribbling speed, shooting accuracy and passing accuracy. Multiple linear regression analysis was used to estimate the relative contributions of age, stage of sexual maturity, height, body mass and years of formal training in soccer to the six skill tests. Age, experience, body size and stage of puberty contributed significantly but in different combinations to the variance in four of the six skill tests: dribbling with a pass (21%; age, stage of maturity), ball control with the head (14%; stage of maturity, height, body height 6 body mass interaction), ball control with the body (13%; stage of maturity, years of training) and shooting accuracy (8%; stage of maturity, height; borderline significance, P = 0.06). There were no significant predictors for the tests of dribbling speed and passing accuracy. In conclusion, age, experience, body size and stage of puberty contributed relatively little to variation in performance in four of the six soccer-specific skill tests in adolescent footballers aged 13 15 years. Keywords: Adolescence, motor performance, puberty, sport-specific skills Introduction Techniques, or sport-specific technical skills, are a central component in the development of young athletes in many sports, including soccer. A variety of tests has been developed to evaluate ball control with the body (trapping), head (heading) and feet (dribbling), passing (short and long), shooting accuracy, throwing and kicking for distance, agility and volleying, among others (Caicedo et al., 1993; Van Rossum & Wijbenga, 1993; Soares et al., 1994; Rösch et al., 2000; see also Kirkendall et al., 1987). The focus is often on the validity of the tests, changes in performance with age among youth players, comparisons of skills in youth and professional players classified by level of competition or expertise, and occasionally relationships between skill tests and outcomes of match-play. Several studies have considered the kinematics of specific skills (Luhtanen, 1988) and the strength and flexibility (Ôberg et al., 1984; Poulmedis, 1985; Togari et al., 1988; Rösch et al., 2000) of players of different skill levels in soccer, and the physical fitness and soccer skills of youth players and non-players (Seabra et al., 2001). The effect of specific training programmes on sportspecific skills, including jumping (Gauffin et al., 1989) and kicking (De Proft et al., 1988), has also received consideration. Studies of youth soccer players have not systematically considered the potential influence of variation in growth and biological maturity status upon sport-specific football skills. Youth soccer players classified as elite and non-elite, or as being high and low in football ability, for example, differ in body size and maturity (Malina, 2003), and in strength, flexibility and soccer-specific skills (Hansen Correspondence: Robert M. Malina, Route 2 Box 140, Bay City, TX 77414, USA. E-mail: rmalina@wcnet.net ISSN 0264-0414 print/issn 1466-447X online ª 2005 Taylor & Francis Group Ltd DOI: 10.1080/02640410410001729928

516 R. H. Malina et al. et al., 1999; Rösch et al., 2000). Unfortunately, individual differences in size and maturity are not considered in the comparisons of such select samples. This study has two purposes: (1) to compare performances on six soccer-specific skill tests by position and stage of sexual maturity in youth players aged 13 15 years of age; and (2) to estimate the contributions of age, experience in the sport, body size and sexual maturity status to performances on six soccer-specific skill tests. Methods Participants The sample included 69 youth soccer players born either in 1984 (n = 39) or 1985 (n = 30). The players were members of teams from three clubs in the Porto area of Portugal. The teams competed in the highest division for their age group in the first Portuguese national division. Within the structure of youth soccer programmes in Portugal, the players were classified as novices (iniciados) at the time of the study. The older players (those born in 1984) will move to the next competitive category, juveniles (juvenis), in the first Portuguese national division, while the younger players (those born in 1985) will remain as novices. Subsequent soccer categories are juniors (juniores or cadetes) and professionals. Measurements and tests The players were measured and tested within a 2- week period, 13 28 February 1999. A decimal age for each player was calculated from his date of birth and the midpoint of the interval during which the measurements and tests were administered (20 February). The ages of the participants ranged from 13.2 to 15.1 years. Height and body mass were measured with standard techniques to the nearest 0.1 cm and 0.1 kg, respectively. Stage of pubic hair based on the criteria of Tanner (1962) was also evaluated at clinical examination by a paediatrician experienced in the assessment of secondary sex characteristics. The players were interviewed to estimate the number of years of regular training in soccer. Six tests of technical skills in soccer were administered (Federação Portuguesa de Futebol, 1986). The tests are recommended by the Federation and are traditionally used in Portugal (e.g. Seabra et al., 2001; Coelho e Silva et al., 2004). The tests were administered outdoors on a playing field. The players warmed-up in the usual manner before a practice session (stretching and jogging), and also rested between tests. The tests were administered in a station format and not in any specific order. Ball control with the body (skill 1) Within a 969 m square, the player had to keep the ball in the air without using the arms or hands. The score recorded was the number of hits of the ball before it fell to the floor. Counting stopped when the ball hit the floor, the participant moved out of the square or he touched the ball with the arms or hands. One trial was administered, although the participant could start the trial again if he failed to contact the ball three times in the initial attempt. Ball control with the head (skill 2) Within a 969 m square, the player had to keep the ball in the air using only the head. The score recorded was the number of hits of the ball before it fell to the floor. Counting stopped when the ball hit the floor, the participant moved out of the square or he touched the ball with any part of the body except the head. One trial was administered, although the participant was allowed to start the trial again if he failed to contact the ball three times in the initial attempt. Dribbling with a pass (speed and accuracy; skill 3) Four cones were placed in a line, 2.25 m apart, within the 969 m square, and a fifth mark, a flat surface such as a bench 1.2 m wide, was placed on the end line. The participant was instructed to dribble the ball around the first four marks in slalom fashion, make a pass to the fifth mark and receive/ control the ball, and dribble around the four marks back to the starting line. The objective was to complete the drill in the fastest time possible without knocking down the cones and without stepping out of the square, controlling the ball only with the feet. If a cone (mark) was knocked over, the participant had to place it upright and continue the test. Two stopwatches were activated by the timers at the starting signal and were stopped when the participant crossed the starting line. The average of the two values was used in the analysis. Dribbling (speed; skill 4) A cone was placed on each corner of the 969 m square (four cones). A fifth cone was place midway (4.5 m) on the line of the square where the test began Thus, the near end had three cones (one on each corner and the third midway) and the far end had two cones (one at each corner). Beginning at one corner, the athlete had to conduct the ball with the feet (dribble) around the three cones (corner directly opposite the starting cone, the cone placed midway, and the cone diagonally opposite the starting cone) in slalom fashion, and then dribble the ball into the fifth cone (i.e. not with a pass). The objective was to complete the drill in the fastest time possible by controlling the ball only with the feet without knocking down the cones. If a cone was knocked over, the participant

Maturity-associated variation in sport-specific skills 517 had to place it upright and continue the test. The overall slalom distance was thus about 40 m. Two stopwatches were activated by the timers at the starting signal and were stopped when the ball was dribbled into the fifth cone. The average of the two values was used in the analysis. Passing (skill 5) Five targets were placed 2.5 m apart at the end line of the 969 m square. The athlete was standing outside of the square at the opposite line of the target. Two attempts at each target were allowed for a total of 10 attempts. The objective was to hit the targets with the kicked ball in succession from one to five; two attempts were permitted for each target. The score was the number of successful target hits; the maximum score was 10 points. Shooting (skill 6) A263m goal was set up at the end line of a 969 m square. The target was divided by ropes into six sections. One rope was placed horizontally between the posts at a height of 1.5 m. Two ropes were dropped from the crossbar, 0.5 m from each post. Five points were allocated for the upper right and left sections, and two points for the upper middle section. Three points were allocated for the lower right and left sections, and one point for the lower middle section. While standing outside of the square at the opposite line of the goal, the player had five attempts at kicking the ball into the goal. The maximum score was 25 points. The reliability of the skills tests was assessed using repeated-measures analysis of variance (ANOVA) and intraclass correlation. The intraclass correlations were 0.70 (passing [skill 5], shooting [skill 6]), 0.71 (dribbling speed [skill 4]), 0.76 (ball control with the head [skill 1]), 0.78 (ball control with the body [skill 2]) and 0.81 (dribbling with a pass [skill 3]). The coefficients indicate reasonable reliability. Since the tests assess important elements of soccer (i.e. ball control, dribbling, passing and shooting), face validity can be assumed. Validity is also suggested in moderate correlations between the six tests of the Federação Portuguesa de Futebol (1986) and soccer tests involving slalom dribbling and wall volleying (Coelho e Silva et al., 2004). Validity coefficients ranged from 0.53 to 0.94 between performances on the slalom dribble and wall volley and judges ratings of soccer playing ability (Kirkendall et al., 1987). Statistical analysis Descriptive statistics for each skill test were calculated for the total sample (n = 69), by stage of puberty (PH1 through PH5) and by position forward/striker (n = 10), midfielder (n = 30) and defender (n = 29). The body size and pubertal characteristics of the players have been described previously (Malina et al., 2004). Differences in the six soccer-specific skills among players by position and maturity status were initially evaluated by ANOVA and Tukey s least significant difference post-hoc test. Given the association among age, body height and body mass, and the association between body size and performance in adolescent males, maturity-associated variation in the six skill tests was assessed using analysis of covariance (ANCOVA), with age, body height and body mass as the covariates. Multiple linear regression analysis was used to estimate the relative contribution of chronological age, stage of sexual maturity, height, body mass and years of formal training in soccer to variation in performances on the six skill tests. Although it has been suggested and seems logical that older youths would have more years of training (Stager et al., 1990), there is no relationship between age and years of training in this sample of adolescent soccer players (r = 0.02). Since height and body mass are highly related, residuals (individual values minus the mean) were used in the regression. A height 6 body mass interaction term was calculated as the product of the residuals for height and body mass. This method reduces the collinearity among the independent variables, making them more stable predictors of the dependent variables (i.e. performance on the six skill tests). The regression protocol permitted all variables to be entered into the equation and then sequentially remove the variables that met the criterion for elimination (backward elimination). In this procedure, the variable with the smallest partial correlation with the dependent variable was considered first for removal; if it met the criterion for removal (P 4 0.10), it was removed. The procedure was repeated for the other variables until those variables that did not meet the removal criterion remained. An alpha level of 0.05 was used in all analyses. Results Descriptive statistics for age and the six skill tests in the total sample and in players grouped by position are summarized in Tables I and II, respectively. There is considerable variation in the scores on the skill tests, but none of the differences in the players grouped by position are significant. Descriptive statistics for the players grouped by stage of pubic hair (PH) are summarized in Table III. Although the number of players in PH1 is small, they are by definition pre-pubertal and biologically different from players in stages PH2 through PH5. Differences in skill scores among maturity categories are not significant, with one exception. Performance

518 R. H. Malina et al. Table I. Descriptive statistics for age and the six skill tests for the total sample (n = 69) of adolescent soccer players Variable Mean s Median Minimum Maximum Age (years) 14.3 0.6 14.3 13.2 15.1 Training (years) 4.5 1.8 5 1 8 Ball control, body (number of hits) 56.5 46.1 42 6 217 Ball control, head (number of hits) 10.5 7.5 7 3 37 Dribbling, speed with pass (s) 8.1 1.1 7.8 5.9 12.6 Dribbling, speed (s) 14.0 1.2 13.9 11.6 19.8 Passing (points) 2.5 1.6 2 0 7 Shooting (points) 7.0 3.2 7 1 14 Table II. Descriptive statistics for age and the six skill tests in adolescent soccer players by position and results of analyses of variance (mean + s) Variable Defence (n = 29) Midfield (n = 30) Forward (n = 10) F Age (years) 14.2 0.6 14.2 0.6 14.7 0.5 3.02 Training (years) 4.2 1.9 4.7 1.4 4.8 2.6 0.81 Ball control, body (number of hits) 55.3 42.4 61.9 52.9 43.7 34.0 0.59 Ball control, head (number of hits) 11.0 7.7 10.8 8.1 8.3 4.7 0.49 Dribbling, speed with pass (s) 8.2 1.4 8.0 0.9 8.1 0.7 0.20 Dribbling, speed (s) 14.1 1.1 14.1 1.5 13.6 0.4 0.75 Passing (points) 2.4 1.3 2.4 1.9 3.0 1.3 0.59 Shooting (points) 6.8 2.8 7.2 3.8 6.6 2.2 0.17 on the dribbling task requiring speed, passing and receiving a pass (skill 3) differs significantly (P50.05) between players in PH2, PH4 and PH5 compared to players in PH1 that is, the players in PH2, PH4 and PH5 completed the drill in a faster time than players in PH1 but did not differ among themselves. All other pairwise comparisons for this dribbling task are not significant. The skill test scores of the players grouped by stage of pubic hair were also compared using ANCOVA, with age, height and body mass as the covariates (Table IV). The differences among maturity categories are not significant with the exception of the dribbling task requiring speed, passing and receiving a pass (skill 3). Only the difference between players in PH1 and PH4 is significant (P 5 0.05) after age, height and body mass are statistically controlled; all other pairwise comparisons are not significant. Estimates of the relative contribution of age, stage of sexual maturity, height, body mass and years of training in soccer to performances on the six skill tests are summarized in Table V. A small but significant percentage of the variance in the two ball control tasks is explained by the independent variables. Stage of maturity and years of training are significant predictors of overall ball control with the body (skill 1), accounting for 13% of the variance (P50.05), and stage of maturity, height and the height 6 body mass interaction are significant predictors of ball control with the head (skill 2), accounting for 14% of the variance (P50.05). It is perhaps of interest that the influence of height on ball control with the head is negative, whereas the influence of the height 6 body mass interaction is positive. Stage of maturity contributes positively to both ball control tasks. Age and stage of maturity are significant predictors of performance in the test requiring dribbling speed, passing and receiving the pass (skill 3), accounting for 21% of the variance (P50.01). Both age and maturity contribute positively to performance on this task. The contribution of stage of maturity (positive) and height (negative) to variation in shooting accuracy (skill 6) is of borderline significance (P = 0.06) and low (8%). In contrast, there are no significant predictors for the tests of dribbling speed (skill 4) and passing accuracy (skill 5). Discussion The means and standard deviations of performances of the present sample of players on the six soccer skill tests are quite similar to those for other samples of elite youth players of the same age range in Portugal (Seabra et al., 2001). Age, years of experience in soccer, body size and maturity status contributed significantly to variation in performance on four soccer-specific skill tests, but the variance explained was relatively small (8 21%) and of borderline significance for the test of shooting accuracy (Table V). On the other hand, the predictor variables did not contribute significantly to performance on two tests (dribbling speed, passing). In regression analysis, standardization allows comparison of the

Maturity-associated variation in sport-specific skills 519 Table III. Descriptive statistics for age and the six skill tests in adolescent soccer players by stage of pubic hair and results of analyses of variance (mean+s) Stages of pubic hair development Variable PH1 (n = 6) PH2 (n = 10) PH3 (n = 13) PH4 (n = 21) PH5 (n = 19) F Age (years) 14.1 0.7 14.3 0.7 14.1 0.6 14.2 0.5 14.5 0.6 0.97 Ball control, body (number of hits) 28.5 12.9 45.8 27.9 41.3 39.8 68.8 45.9 68.6 57.6 1.95 Ball control, head (number of hits) 8.8 5.5 12.2 9.0 7.4 4.0 9.7 7.1 13.1 8.9 1.46 Dribbling, speed with pass (s) 9.2 1.2 8.1 0.8 8.5 1.4 7.8 0.9 7.7 1.0 3.41* Dribbling, speed (s) 14.7 1.0 14.2 2.0 13.8 0.9 14.1 1.2 13.8 1.0 0.79 Passing (points) 2.3 2.6 2.2 1.3 2.0 1.2 2.9 1.6 2.6 1.6 0.70 Shooting (points) 6.8 4.2 6.9 3.3 6.2 3.6 7.5 3.3 6.9 2.5 0.30 * P = 0.01; the post-hoc test indicated that players in PH2, PH4 and PH5 differed significantly (P 5 0.05) from players in PH1. Other pairwise comparisons were not significant. Table IV. Adjusted means and standard errors for the six skill tests in adolescent soccer players by stage of pubic hair based on ANCOVA, with age, height and body mass as the covariates Stages of pubic hair development Variable PH 1 PH 2 PH 3 PH 4 PH 5 F Ball control, body (number of hits) 36.0 22.0 50.5 16.4 43.1 12.6 70.2 9.9 60.1 13.0 1.53 Ball control, head (number of hits) 4.8 3.6 9.3 2.7 7.9 2.0 9.7 1.6 15.6 2.1 1.60 Dribbling, speed with pass (s) 9.0 0.5 8.0 0.4 8.4 0.3 7.8 0.2 7.9 0.3 3.05* Dribbling, speed (s) 15.4 0.6 14.6 0.5 13.8 0.4 14.1 0.3 13.4 0.4 1.01 Passing (points) 2.5 0.8 2.2 0.6 2.0 0.5 2.8 0.4 2.6 0.5 0.48 Shooting (points) 4.9 1.6 5.6 1.2 6.3 0.9 7.5 0.7 8.2 0.9 0.94 * P50.01; the post-hoc test indicated that only the difference between players in PH1 and PH4 was significant (P 5 0.05). Other pairwise comparisons were not significant. Table V. Predictors of performance on the skill tests and estimated R 2 in adolescent soccer players based on multiple regression analyses Adjusted Performance task Predictor Standardized beta coefficients R 2 R 2 P Ball control, body Maturity 0.308 Training 0.200 0.13 0.10 5 0.05 Ball control, head Maturity 0.430 Height 70.340 Height 6 mass 0.251 0.14 0.10 5 0.05 Dribble, speed with pass Age 0.295 a Maturity 0.312 a 0.21 0.19 5 0.01 Dribble, speed No significant predictors Passing No significant predictors Shooting Maturity 0.317 Height 70.391 0.08 0.05 0.06 Note: The adjusted R 2 allows for the degrees of freedom associated with the sum of squares and is generally considered as a more accurate measure of goodness-of-fit than R 2. a The signs of the coefficients were reversed because a lower time is a better performance. estimated contribution of each variable using the coefficients. They are not related to the scale of the raw data, and as such should be interpreted without scale. Positive beta coefficients suggest an increase, while negative coefficients suggest a decrease, in performance associated with changes in particular independent variables. Biological maturity status contributes positively, albeit to a small extent, to the variation in four of the tests: ball control with the body (skill 1), ball control with the head (skill 2), dribbling speed with a pass for accuracy at the end of the test (skill 3), and shooting accuracy (skill 6). This suggests that within this relatively homogeneous sample of adolescent soccer

520 R. H. Malina et al. players, advanced biological maturity status is associated with slightly better performance on four of the six skill tests. The beta coefficients for height are negative for two of the skill tests, ball control with the head (skill 2) and shooting accuracy (skill 6). This suggests that shorter height is associated with slightly better performance on these two tests. Body mass does not appear as a significant source of variation in any of the skill tests, but appears as a significant predictor only in the height 6 body mass interaction for ball control with the head (skill 2). Chronological age and years of training in soccer contribute significantly to two of the skill tests, dribbling speed with a pass (skill 3) and ball control with the body (skill 1), respectively (Table V). Note, however, that age and years of experience are not related in this sample of soccer players (r = 0.02). This probably reflects the relatively narrow distribution of ages of the players (13.2 15.1 years). The influence of age and training on the two skill tests is complemented by stage of maturity. This suggests that older, more mature players perform slightly better in the dribbling test, and that more mature players with more experience in soccer perform slightly better in ball control with the body. In this sample of 13- to 15-year-old soccer players, pubertal status, years of training and body size accounted for a greater percentage of the variance in three tests of functional capacity (aerobic resistance, speed, power), but in different combinations (Malina et al., 2004). Maturity and years of experience accounted for 21% of the variance in aerobic resistance; body mass and maturity accounted for 50% of the variance in running speed (30 m dash); and body height and maturity accounted for 41% of the variance in explosive power (vertical jump). Thus, body size and maturity impact performance on tests of functional capacity, specifically running speed and explosive power, more so than soccer-specific skills. Variation in performances on the six soccer skills tests associated with age differs among the tests in Portuguese youth players aged 10 16 years (Seabra et al., 2001). The players were grouped into three playing categories, infantis (11.7 + 0.4 years), iniciados (13.5 + 0.5 years) and juvenis (16.1 + 0.5 years). Comparisons of the three groups showed, on average, improved performances with age in the two ball control tests (skills 1 and 2), but no major differences between age groups in the other four tests. The influence of body size and stage of puberty on performances on the soccer skills tests was not considered. In one of the more comprehensive studies of soccer-specific skills, Rösch et al. (2000; see also Junge et al., 2000) compared a combined sample of youth players from Germany, France and the Czech Republic in two age groups, 14 16 years (14.4 16.8) and 16 18 years (15.6 18.9), and at two levels (high, low) within each age group. In the younger sample (14 16 years), players classified as high level were, on average, older by about 6 months, taller and heavier. They also performed significantly better than those classified as low level on five juggling (ball control) tests and tests of dribbling speed and long and short passing. The high level players also performed significantly better in functional tests related to speed (sprints), power (jumps) and aerobic capacity (distance run). Unfortunately, the chronological age difference between the samples was not controlled in making the statistical comparisons, and the contributions of age and body size to performances were not considered. In the older age group (16 18 years), age and body size differences between high and low level players were reduced and not significant (Junge et al., 2000). This reflects the catch-up in body size as males approach maturity in late adolescence. Note, however, that the maturity status of the youth players was not considered in this comprehensive study. On the other hand, differences in mean performances of high and low level players were slightly smaller, though generally significant, on the soccer-specific skill and functional capacity tests (Rösch et al., 2000). The results of other studies of sport-specific skill and functional capacity tests in youth soccer players are generally consistent with the comprehensive study of Rösch et al. (2000). These include comparisons of the strength of elite and non-elite 11-year-old Danish players (Hansen et al., 1999) and comparisons of the shuttle run and vertical jump performances of elite (national team) and non-elite (regional teams) late adolescent Portuguese players (Garganta et al., 1993). The elite players performed better. The elite 11-year-old players were also, on average, slightly older, advanced in biological maturity status and had a larger body size. Statistical adjustment for age and body dimensions reduced the differences between elite and non-elite players, but the significant influence of maturity status on strength persisted (Hansen et al., 1999). The elite late adolescent players were also older, taller and heavier than the non-elite players. The comparison did not consider the age and size differences (an indicator of biological maturity status was not included in the study) between elite and non-elite players in making the comparisons (Garganta et al., 1993). Overall, the contributions of age, years of experience, body size and maturity status to performances in four of the six soccer-specific skill tests are

Maturity-associated variation in sport-specific skills 521 relatively small (8 21%). Stage of puberty makes a significant but small contribution to variation in the four tests, which suggests that within this relatively homogeneous sample of adolescent football players, advanced biological maturity status is associated with slightly better technical performance. Except for static strength, chronological age and skeletal age alone, or in combination with height and body mass, explained relatively small percentages of variation (0 17%) in several motor performance tasks of adolescent boys (Beunen et al., 1981). However, the largest percentage of variation in performance accounted for by these independent variables generally occurred at 14 and 15 years of age. The same is suggested in comparisons of adolescent boys of contrasting maturity status. Boys advanced in maturity perform, on average, better in agility (shuttle run), speed of limb movement (plate tapping) and vertical jump tests, among others, and the differences are most apparent between 14 and 16 years (Malina et al., 2004). Nevertheless, most of the variation in skill test performance is not accounted for by the predictor variables. This comes as no surprise. Skill is more difficult to measure than physiological indicators such as speed and power. Many factors other than body size and maturity status (e.g. neural control of movement and perceptual-cognitive skills such as anticipation and visual search strategies) influence performances on sport-specific skill tests (see Williams and Franks, 1998; Williams, 2000; Williams and Reilly, 2000). Maturation of neural control and perceptual-cognitive skills during adolescence is another source of variation. Other factors that need to be considered include coach athlete interactions and individual differences in responsiveness to instruction and practice. The results of the present study highlight the need to extend studies of soccerspecific skills to include other potential determinants. Acknowledgement The assistance of Manuel Coelho e Silva and Antonio Figueiredo of the Faculty of Sport Sciences and Physical Education, University of Coimbra, is acknowledged and gratefully appreciated. References Beunen, G., Ostyn, M., Simons, J., Renson, R., & Van Gerven, D. (1981). Chronological and biological age as related to physical fitness in boys 12 to 19 years. Annals of Human Biology, 8, 321 331. Caicedo, J.G., Matsudo, S.M.M., & Matsudo, V.K.R. (1993). Teste especifico para mensurar agilidade em futebolistas e sua correlação com o desempenho no passe im situação real de jogo. Revista Brasileira de Ciencia e Movimento, 7(2), 7 15. Coelho e Silva, M., Figueiredo, A., & Malina, R.M. (2004). Avaliação da mestria no inicio da preparação de jovens futebolistas. Horizonte: Revista de Educaçâo Fisica e Desporto, 19, 23 32. De Proft, E., Cabri, J., Dufour, W., & Clarys, J.P. (1988). Strength training and kick performance in soccer players. In Science and Football (edited by T. Reilly, A. Lees, K. Davids and W.J. Murphy), pp. 109 113. London: E & FN Spon. Federação Portuguesa de Futebol (1986). Habilidades e destrezas do futebol: Os skills do futebol. Lisbon: Editora Federação Portuguesa de Futebol. Garganta, J., Maia, J., Silva, R., & Natal, A. (1993). A comparative study of explosive leg strength in elite and non-elite young soccer players. In Science and Football II (edited by T. Reilly, J. Clarys and A. Stibbe), pp. 304 306. London: E & FN Spon. Gauffin, H., Ekstrand, J., Arnesson, L., & Tropp, H. (1989). Vertical jump performance in soccer players: a comparative study of two training programs. Journal of Human Movement Studies, 16, 215 224. Hansen, L., Bangsbo, J., Twisk, J., & Klausen, K. (1999). Development of muscle strength in relation to training and level of testosterone in young male soccer players. Journal of Applied Physiology, 87, 1141 1147. Junge, A., Dvorak, J., Chomiak, J., Peterson, L., & Graf- Baumann, T. (2000). Medical history and physical findings in football players of different ages and skill levels. American Journal of Sports Medicine, 28(suppl.), S16 S21. Kirkendall, D.R., Gruber, J.J., & Johnson, R.E. (1987). Measurement and Evaluation for Physical Educators. Champaign, IL: Human Kinetics. Luhtanen, P. (1988). Kinematics and kinetics of maximal instep kicking in junior soccer players. In Science and Football (edited by T. Reilly, A. Lees, K. Davids and W.J. Murphy), pp. 441 448. London: E & FN Spon. Malina, R.M. (2003). Growth and maturity status of young soccer (football) players. In Science and Soccer, 2nd edn. (edited by T. Reilly and M. Williams), pp. 287 306. London: Routledge. Malina, R.M., Bouchard, C., Bar-Or, O. (2004). Growth, Maturation, and Physical Activity, 2nd edn. Champaign, IL: Human Kinetics. Malina, R.M., Eisenmann, J.C., Cumming, S.P., Ribeiro, B., & Aroso, J. (2004). Maturity-associated variation in the growth and functional capacities of elite youth football (soccer) players 13 15 years. European Journal of Applied Physiology, 91, 555 562. Ôberg, B., Ekstrand, J., Möller, M., & Gillquist, J. (1984). Muscle strength and flexibility in different positions of soccer players. International Journal of Sports Medicine, 5, 213 216. Poulmedis, P. (1985). Isokinetic maximal torque power of Greek elite soccer players. Journal of Orthopedics and Sports Physical Therapy, 6, 293 295. Rösch, D., Hodgson, R., Peterson, L. et al. (2000). Assessment and evaluation of football performance. American Journal of Sports Medicine, 28(suppl), S29 S39. Seabra, A., Maia, J.A., & Garganta, R. (2001). Crescimento, maturação, aptidão fisica, força explosiva e habilidades motoras especificas. Estudo em jovens futebolistas e não futebolistas do sexo masculino dos 12 aos 16 anos de idade. Revista Portuguesa de Ciencias do Desporto, 1(2), 22 35. Soares, E.A., Matsudo, V.K.R., Ferreira, M., Figueira, A.J., & Leandro de Araujo, T. (1994). Relationship among agility tests in junior high soccer players. In Proceedings of the XIX Simposio Internacional de Ciencias do Esporte, p. 135. Sao Caetano, Sao Paulo: Centro de Estudos do Laboratorio de Aptidao Fisica de Sao Caetano do Sul. Stager, J.M., Wigglesworth, J.K., & Hatler, L.K. (1990). Interpreting the relationship between age of menarche and prepubertal training. Medicine and Science in Sports and Exercise, 22, 54 58.

522 R. H. Malina et al. Tanner, J.M. (1962). Growth at Adolescence, 2nd edn. Oxford: Blackwell. Togari, N., Ohashi, J., & Ohgushi, T. (1988). Isokinetic muscle strength of soccer players. In Science and Football (edited by T. Reilly, A. Lees, K. Davids and W.J. Murphy), pp. 181 185. London: T & FN Spon. Van Rossum, J.H.A., & Wijbenga, D. (1993). Soccer skills technique tests for youth players: construction and implications. In Science and Football II (edited by T. Reilly, J. Clarys and A. Stibbe), pp. 313 318. London: E & FN Spon. Williams, A.M. (2000). Perceptual skill in soccer: implications for talent identification and development. Journal of Sports Sciences, 18, 737 750. Williams, A.M., & Franks, A. (1998). Talent identification in soccer. Sports, Exercise and Injury, 4, 159 165. Williams, A.M., & Reilly, T. (2000). Talent identification and development in soccer. Journal of Sports Sciences, 18, 657 667.