Perceptual ond Motor Skills, 1972, 75, 211-217.
@ Perceptual and Motor Skills 1992
CATCHING ACTION: VISUOMOTOR ADAPTATIONS I N CHILDREN '" J O H N G. WILLIAMS
The University of Liuerpool Summary.-Videotapes of the catching action of 2 8 children aged 4 to 10 years were carefully observed from simultaneous frontal and lateral perspectives. For both one- and two-handed ball catching, three discernible modes of visual attention and limb movement spaced along a maturity continuum were determined. Only for the ten-year-olds catching two-handed did strategies of visual attention (predicting where the ball would be) and limb usage (grasping with the fingers) couple exclusively as in skilled, adult catching. Success in two-handed catching improved exponentially with age from 77% to 96%. For one-handed catching the success rate was 40% at ages 4, 5, and 6 years, 7.5 and 30% at 7 and 8 years, improving steadily to 92% 3t 10 years. In the middle of age range the drop in performance coincided with the l g h e s t incidence of mixed strategies. No gender differences on either strategy or performance were evident.
Catching skill develops from an early age. Infants act in a manner which has been interpreted as "catching" during the first six months of life (2). At this age children can be observed to "acquire" visually then reach toward, grasp, and retain attractive objects that are presented within their personal space ( 3 , 4). This they achieve by reaching and "scooping" the object toward the body (frequently the mouth!) first with one hand then the other to assist in securing it (10). Trevarthen (8) studied slow-motion film of children from the first week of life until age 6 months and described an erratic pattern of reach and grasp adapting to become quite controlled by the end of the fourth month. Hofsten (2), using similar techniques, provided descriptions and kinematic data of catching by children under one year of age. The actions described and acceleration profiles of the hand as it "homes" onto the target generally resemble those reported in studies of reaching and grasping common objects (5) and ball catching by mature subjects (1). Although the substrate for catching objects is in place very early in life, the difference between an infant "catching7' an attractive object which is fixed to the source of its motion and an object in free flight is considerable. Some two or more years of development and experience must elapse before a child can attempt to "catch" a ball projected toward it by another person. 'Thanks ate extended to the teachers and pupils of the Millwood Primary School, Speke, Merseyside for their assistance with this project. The contributions and critical observations of Steve Beevers, Pauline Phillips, Martin Turner, and John Buckles were very helpful. Address enquiries to Dr. John Williams, Department of Movement Science and Physical Education, The University of Liverpool, Liverpool L69 3BX, England.
212
J. G . WILLIAMS
Even when a child has achieved sound control of upright posture, is able to walk steadily and manipulate various objects in its environment, the level of skill required to participate in simple catching games normally takes at least six to eight years to acquire. Research on "real" ball catching has focussed on older children (7, 11) and already skilled, adult subjects ( I , 9). The exception to this is a study by Kay (6) who described the actions of three children aged 2, 5 , and 15 years while catching a ball thrown to them from a distance of three feet. Film analysis showed that the two-year-old child held its hands together palms up, its eyes watched the thrower or the thrower's hands, the ball fell into its arms with no closure of the hands. Marked change in strategy was observed in the five-year-old. This child watched the ball through the entire flight, moved its hands to meet the ball with fingers cupped so as to retain it. However, there was little anticipation at this age. By contrast, the 15-yearold moved to intercept the ball and grasped with the fingers. A clear developmental change was that visual attention at the youngest age centred mainly on the origin of the ball's motion but not the catcher's hands or the ball, on both the ball and the catcher's own hands at the middle age, and predominantly on the ball by the eldest child. Kay's study provided a useful foundation for the description of the ballcatching action but it is plain that more comprehensive detail which maps the gradual transition from infant to adult behaviour would be a necessary start to understand the visuomotor processes involved and any critical developmental periods. This provided the motive for the present study. The primary purpose was to record, carefully observe, and describe adaptations in the way children between the ages of four and ten years used their visual attention and their limbs to catch a ball which was thrown to them by another person. A subsidiary and related purpose was to make similar observations of the same children when they were constrained to use only the preferred hand. The one-handed catch was included because the abdity to deploy one's limbs adaptively and spontaneously to complete a motor task successfully is a mark of higher proficiency. I n certain circumstances, the accomplishment of catching with one hand rather than being constrained to two exemplifies greater skillfulness. The manner in which children cope with such a constraint is intrinsically interesting and could signal critical phases in the development of a skiU when a particular type of practice was appropriate or necessary. Under both conditions, success in preventing the ball from falling to the ground was also recorded without judgement of the quality of catch.
METHOD Subjects Twenty-eight children aged between 4 and 10 years were selected at ran-
CHILDREN'S CATCHING ACTION
2 13
dom by their teachers from the infant and junior classes of a large primary school in the Merseyside area of Northwest England. There were two boys and two girls from similar socioeconomic background in each of seven age bands. he children participated with their own and their parents' consent. All subjects had normal, uncorrected vision and no history of perceptual or motor problems.
Procedure The catching trials were administered individually. Each child came with their class teacher to a specially prepared area of the school hall where they were met by a physical education teacher who had taught them for several lessons. This person explained that the child would be asked to stand in a specified area facing her. She would stand a short distance away and throw a ball toward the child who would try to catch the ball. After retaining the ball for a few seconds the child was to throw it back to the teacher. The child would do this 20 times using both hands and, after a short rest, a further 20 times using one hand only. Preliminary study indicated that randomization was inadvisable. Each child chose a PVC-coated ball from a selection of different sizes (60, 70, and 80 mm diameter: all subjects chose 70 mm). Size of ball was checked against the dimensions of the child's hand. It was ensured that the ball matched the likely ability of the child and that the child was confident of being able to catch in the manner required. A pilot study showed that children aged 6 and under preferred a 2-metre distance and those aged 7 and above 3 metres plus. The ball-flight distance was set at 2 metres for the 4-, 5-, and 6-year-olds and 3 metres for the older age groups. Also, this study indicated children's preference for a human rather than mechanical ball-projection system. An underarm toss of the ball was aimed at the torso at about shoulder height. Each ball toss was independently judged by an observer at the time of delivery. This person performed a visual check of the peak height and lateral deviation of the ball's path relative to the top of the catcher's head and shoulder extremities on a videomonitor which displayed a frontal view of the catcher. Further trials were administered when clear trajectory and lateral deviations were apparent. Every catch was recorded on videotape using cameras positioned at lateral and frontal viewpoints. The lateral camera distance and lens focus was adjusted to capture both the ball's flight and the catcher's action. The zoom lens of the frontal camera was adjusted to provide a clear view of the subjects' eyes and line of sight relative to the approaching ball. The consistency of each delivery was checked using videoplayback on a h g h quality monitor. This showed minimal lateral deviation, average velocity of 2 m/sec. ( + 0.03), and constant trajectory.
214
J. G. WILLIAMS
Classification of Visual Attention and Movement Strategies Videotaped copies of all catching trials were given to four experienced physical education teachers who had recently undertaken a graduate course in motor skill analysis with an emphasis on videoanalysis techniques. They were asked to observe the actions of each child on each catching trial using the slow-motion and frame-search functions of a high quality, industrial standard videotape recorder and monitor. The author requested that they attempt to characterise and record both the visual attention and limb movements of the subjects as they attempted to catch the ball. The observers were specifically asked to examine and describe the frontal view of the catcher throughout the epoch of the catch and to ascertain the subjects' line of sight relative to the thrower and the ball and also to examine and describe the movements of the limbs during the epoch of the catch. When this task was complete, the observers met with the author and reported their findings. Following several further sessions of viewing the videotapes as a group, full agreement was reached that both types of catch could be classified into three broadly independent "categories" of attentional and movement behaviour. Visual attention.-(a) The catcher looks toward the thrower and particularly that person's hands prior to the release of the ball. Attention remains fixed at the point at which the throw originated whilst the ball is in flight. When the ball's progress is arrested by the catcher, attention changes to confirm that the object has landed where expected. The over-all impression is that the catcher's attention lags behind events. This strategy was labelled retrospective. (b) The catcher looks at the thrower's hands and, as the ball leaves them, the eyes appear to attempt to match the position of the ball throughout its entire flight and do this right into the catching hand(s). This strategy was labelled concurrent. (c) The catcher observed the ball intently during the first few frames (120 to 200 milliseconds). Thereafter, the gaze appeared to be less focussed as if the decision on precisely how to act had been made in that early period of the ball's flight. Even as the ball enters the hand, attention was not specifically directed at either ball or hand but both. The impression was that, as the ball was received, vision confirms that the act was appropriate even though there were occasions when the object was fumbled or dropped. This strategy was labelled predictive. Movement.--(a) The catcher prepares for the b d ' s arrival well before it is projected by cradling the arm(s) to make a receptacle into which it is expected that the ball will fall. The disposition of the arms remains unchanged until the object either falls directly into the receptacle or, more frequently, hits the torso then faUs in. At this point arm(s) and hand($, working as a single unit, flex to retain the ball by trapping it against the body. When permitted to use only one hand, subjects attempted to do the same they did
CHILDREN'S CATCHING ACTION
215
when free to use both hands. This strategy was labelled cradling. (b) The subject reaches toward the oncoming ball with hands placed palm upwards during the first thud of the ball's flight, rotating to a neutral position (palms facing each other) during the middle of the trajectory then closing together in an attempt to "clamp" the ball before it falls to the ground. When attempting to catch with one hand, the early part of the action is similar, the hand moved toward the ball, palm uppermost. I n the latter part of the action, the catcher tries to place the hand underneath the ball as it falls and grasps with palm uppermost to hold it. When this action was unsuccessful, subjects were observed to regress to a strategy which was half cradling as described above and half clamping whereby the manipulative phase of their action was slightly delayed and an attempt was made to trap the ball by flexing the elbow and clutching it to the torso. This strategy was called clamping. (c) The subject reaches toward the oncoming ball with palms of hands facing the ball, fingers flexed ready to admit the ball to the hand and secure it. The ball is intercepted well away from the body and sometimes at the limit of elbow extension. This strategy was termed grasping.
RESULTS AND DISCUSSION The subjects' use of these strategies was assessed by the eight observers. Each had received instruction in motor skill analysis using videotape techniques. First, the purpose of the project and the classification system was explained; examples of both the attentional and manual strategies were shown on a videomonitor in real-time, slow motion, and frame-by-frame until a clear frame-of-reference for their observations was established. Second, each catch was watched several times in the various videoplayback modes and the observers made independent judgements relative to both visual attention and movement. Interjudge reliability was assessed by computing coefficients of concordance for both visual and movement strategies. The values were W = 0.79 and 0.81 (p
@ Perceptual and Motor Skills 1992
CATCHING ACTION: VISUOMOTOR ADAPTATIONS I N CHILDREN '" J O H N G. WILLIAMS
The University of Liuerpool Summary.-Videotapes of the catching action of 2 8 children aged 4 to 10 years were carefully observed from simultaneous frontal and lateral perspectives. For both one- and two-handed ball catching, three discernible modes of visual attention and limb movement spaced along a maturity continuum were determined. Only for the ten-year-olds catching two-handed did strategies of visual attention (predicting where the ball would be) and limb usage (grasping with the fingers) couple exclusively as in skilled, adult catching. Success in two-handed catching improved exponentially with age from 77% to 96%. For one-handed catching the success rate was 40% at ages 4, 5, and 6 years, 7.5 and 30% at 7 and 8 years, improving steadily to 92% 3t 10 years. In the middle of age range the drop in performance coincided with the l g h e s t incidence of mixed strategies. No gender differences on either strategy or performance were evident.
Catching skill develops from an early age. Infants act in a manner which has been interpreted as "catching" during the first six months of life (2). At this age children can be observed to "acquire" visually then reach toward, grasp, and retain attractive objects that are presented within their personal space ( 3 , 4). This they achieve by reaching and "scooping" the object toward the body (frequently the mouth!) first with one hand then the other to assist in securing it (10). Trevarthen (8) studied slow-motion film of children from the first week of life until age 6 months and described an erratic pattern of reach and grasp adapting to become quite controlled by the end of the fourth month. Hofsten (2), using similar techniques, provided descriptions and kinematic data of catching by children under one year of age. The actions described and acceleration profiles of the hand as it "homes" onto the target generally resemble those reported in studies of reaching and grasping common objects (5) and ball catching by mature subjects (1). Although the substrate for catching objects is in place very early in life, the difference between an infant "catching7' an attractive object which is fixed to the source of its motion and an object in free flight is considerable. Some two or more years of development and experience must elapse before a child can attempt to "catch" a ball projected toward it by another person. 'Thanks ate extended to the teachers and pupils of the Millwood Primary School, Speke, Merseyside for their assistance with this project. The contributions and critical observations of Steve Beevers, Pauline Phillips, Martin Turner, and John Buckles were very helpful. Address enquiries to Dr. John Williams, Department of Movement Science and Physical Education, The University of Liverpool, Liverpool L69 3BX, England.
212
J. G . WILLIAMS
Even when a child has achieved sound control of upright posture, is able to walk steadily and manipulate various objects in its environment, the level of skill required to participate in simple catching games normally takes at least six to eight years to acquire. Research on "real" ball catching has focussed on older children (7, 11) and already skilled, adult subjects ( I , 9). The exception to this is a study by Kay (6) who described the actions of three children aged 2, 5 , and 15 years while catching a ball thrown to them from a distance of three feet. Film analysis showed that the two-year-old child held its hands together palms up, its eyes watched the thrower or the thrower's hands, the ball fell into its arms with no closure of the hands. Marked change in strategy was observed in the five-year-old. This child watched the ball through the entire flight, moved its hands to meet the ball with fingers cupped so as to retain it. However, there was little anticipation at this age. By contrast, the 15-yearold moved to intercept the ball and grasped with the fingers. A clear developmental change was that visual attention at the youngest age centred mainly on the origin of the ball's motion but not the catcher's hands or the ball, on both the ball and the catcher's own hands at the middle age, and predominantly on the ball by the eldest child. Kay's study provided a useful foundation for the description of the ballcatching action but it is plain that more comprehensive detail which maps the gradual transition from infant to adult behaviour would be a necessary start to understand the visuomotor processes involved and any critical developmental periods. This provided the motive for the present study. The primary purpose was to record, carefully observe, and describe adaptations in the way children between the ages of four and ten years used their visual attention and their limbs to catch a ball which was thrown to them by another person. A subsidiary and related purpose was to make similar observations of the same children when they were constrained to use only the preferred hand. The one-handed catch was included because the abdity to deploy one's limbs adaptively and spontaneously to complete a motor task successfully is a mark of higher proficiency. I n certain circumstances, the accomplishment of catching with one hand rather than being constrained to two exemplifies greater skillfulness. The manner in which children cope with such a constraint is intrinsically interesting and could signal critical phases in the development of a skiU when a particular type of practice was appropriate or necessary. Under both conditions, success in preventing the ball from falling to the ground was also recorded without judgement of the quality of catch.
METHOD Subjects Twenty-eight children aged between 4 and 10 years were selected at ran-
CHILDREN'S CATCHING ACTION
2 13
dom by their teachers from the infant and junior classes of a large primary school in the Merseyside area of Northwest England. There were two boys and two girls from similar socioeconomic background in each of seven age bands. he children participated with their own and their parents' consent. All subjects had normal, uncorrected vision and no history of perceptual or motor problems.
Procedure The catching trials were administered individually. Each child came with their class teacher to a specially prepared area of the school hall where they were met by a physical education teacher who had taught them for several lessons. This person explained that the child would be asked to stand in a specified area facing her. She would stand a short distance away and throw a ball toward the child who would try to catch the ball. After retaining the ball for a few seconds the child was to throw it back to the teacher. The child would do this 20 times using both hands and, after a short rest, a further 20 times using one hand only. Preliminary study indicated that randomization was inadvisable. Each child chose a PVC-coated ball from a selection of different sizes (60, 70, and 80 mm diameter: all subjects chose 70 mm). Size of ball was checked against the dimensions of the child's hand. It was ensured that the ball matched the likely ability of the child and that the child was confident of being able to catch in the manner required. A pilot study showed that children aged 6 and under preferred a 2-metre distance and those aged 7 and above 3 metres plus. The ball-flight distance was set at 2 metres for the 4-, 5-, and 6-year-olds and 3 metres for the older age groups. Also, this study indicated children's preference for a human rather than mechanical ball-projection system. An underarm toss of the ball was aimed at the torso at about shoulder height. Each ball toss was independently judged by an observer at the time of delivery. This person performed a visual check of the peak height and lateral deviation of the ball's path relative to the top of the catcher's head and shoulder extremities on a videomonitor which displayed a frontal view of the catcher. Further trials were administered when clear trajectory and lateral deviations were apparent. Every catch was recorded on videotape using cameras positioned at lateral and frontal viewpoints. The lateral camera distance and lens focus was adjusted to capture both the ball's flight and the catcher's action. The zoom lens of the frontal camera was adjusted to provide a clear view of the subjects' eyes and line of sight relative to the approaching ball. The consistency of each delivery was checked using videoplayback on a h g h quality monitor. This showed minimal lateral deviation, average velocity of 2 m/sec. ( + 0.03), and constant trajectory.
214
J. G. WILLIAMS
Classification of Visual Attention and Movement Strategies Videotaped copies of all catching trials were given to four experienced physical education teachers who had recently undertaken a graduate course in motor skill analysis with an emphasis on videoanalysis techniques. They were asked to observe the actions of each child on each catching trial using the slow-motion and frame-search functions of a high quality, industrial standard videotape recorder and monitor. The author requested that they attempt to characterise and record both the visual attention and limb movements of the subjects as they attempted to catch the ball. The observers were specifically asked to examine and describe the frontal view of the catcher throughout the epoch of the catch and to ascertain the subjects' line of sight relative to the thrower and the ball and also to examine and describe the movements of the limbs during the epoch of the catch. When this task was complete, the observers met with the author and reported their findings. Following several further sessions of viewing the videotapes as a group, full agreement was reached that both types of catch could be classified into three broadly independent "categories" of attentional and movement behaviour. Visual attention.-(a) The catcher looks toward the thrower and particularly that person's hands prior to the release of the ball. Attention remains fixed at the point at which the throw originated whilst the ball is in flight. When the ball's progress is arrested by the catcher, attention changes to confirm that the object has landed where expected. The over-all impression is that the catcher's attention lags behind events. This strategy was labelled retrospective. (b) The catcher looks at the thrower's hands and, as the ball leaves them, the eyes appear to attempt to match the position of the ball throughout its entire flight and do this right into the catching hand(s). This strategy was labelled concurrent. (c) The catcher observed the ball intently during the first few frames (120 to 200 milliseconds). Thereafter, the gaze appeared to be less focussed as if the decision on precisely how to act had been made in that early period of the ball's flight. Even as the ball enters the hand, attention was not specifically directed at either ball or hand but both. The impression was that, as the ball was received, vision confirms that the act was appropriate even though there were occasions when the object was fumbled or dropped. This strategy was labelled predictive. Movement.--(a) The catcher prepares for the b d ' s arrival well before it is projected by cradling the arm(s) to make a receptacle into which it is expected that the ball will fall. The disposition of the arms remains unchanged until the object either falls directly into the receptacle or, more frequently, hits the torso then faUs in. At this point arm(s) and hand($, working as a single unit, flex to retain the ball by trapping it against the body. When permitted to use only one hand, subjects attempted to do the same they did
CHILDREN'S CATCHING ACTION
215
when free to use both hands. This strategy was labelled cradling. (b) The subject reaches toward the oncoming ball with hands placed palm upwards during the first thud of the ball's flight, rotating to a neutral position (palms facing each other) during the middle of the trajectory then closing together in an attempt to "clamp" the ball before it falls to the ground. When attempting to catch with one hand, the early part of the action is similar, the hand moved toward the ball, palm uppermost. I n the latter part of the action, the catcher tries to place the hand underneath the ball as it falls and grasps with palm uppermost to hold it. When this action was unsuccessful, subjects were observed to regress to a strategy which was half cradling as described above and half clamping whereby the manipulative phase of their action was slightly delayed and an attempt was made to trap the ball by flexing the elbow and clutching it to the torso. This strategy was called clamping. (c) The subject reaches toward the oncoming ball with palms of hands facing the ball, fingers flexed ready to admit the ball to the hand and secure it. The ball is intercepted well away from the body and sometimes at the limit of elbow extension. This strategy was termed grasping.
RESULTS AND DISCUSSION The subjects' use of these strategies was assessed by the eight observers. Each had received instruction in motor skill analysis using videotape techniques. First, the purpose of the project and the classification system was explained; examples of both the attentional and manual strategies were shown on a videomonitor in real-time, slow motion, and frame-by-frame until a clear frame-of-reference for their observations was established. Second, each catch was watched several times in the various videoplayback modes and the observers made independent judgements relative to both visual attention and movement. Interjudge reliability was assessed by computing coefficients of concordance for both visual and movement strategies. The values were W = 0.79 and 0.81 (p
