The influence of verbal kinematic information feedback during the performance of a serial gymnastic skill

Introduction

Irrespective of the frequency of provision, augmented external feedback in the form of knowledge of results (KR), knowledge of performance (KP), or kinematic information feedback (KIF), is considered as one of the most important variables in the learning process of motor skills [1]. Two types of scheduling variables are absolute frequency and relative frequency of feedback. The absolute number of times feedback is given in an instructional progression is referred to as absolute frequency, while relative frequency is the total number of times feedback is given relative to the total number of trials attempted. Several studies have revealed that variations in KR scheduling which reduce the relative frequency of feedback during acquisition prove to be more beneficial for long-term skill retention than practice conditions with feedback provided more often [21, 28]. Additionally, little research has examined the influence of reduced frequency of KP on learning sport skills. Weeks and Kordus [24] examined the effects of variations in relative frequency of KP – 100% (after each trial) and 33% (summary feedback after 3 trials) – on acquisition, retention, and transfer of form for a multilimb closed sport skill. Although no group differences were found for accuracy scores, the 33% group had higher form scores in acquisition, retention and transfer tests.

The notion that in single degree of freedom or simple tasks the provision of lower amount of feedback with KR results in greater amount of learning [19, 25] seems to be verified in complex motor skills, where the great number of degrees of freedom is rapidly increasing.

During the performance of gymnastic skills, different body limbs must be coordinated and acting upon the spatial and temporal restrictions of the skill [11]. In this case, kinematic information feedback related to the limb position was regarded as a basic prerequisite for learning these complex skills. For this reason, the subjects received technical instructions of the pattern of the motion related to the position, the speed or the acceleration of the limbs, and the information related to the actions upon the coordination of these limbs [14, 16]. Because single degree of freedom tasks have been the mainstay of the studies examining the benefits of reduced relative frequency of augmented feedback, the need exists to study variations in feedback scheduled for KIF in motor skills [15] especially in gymnastic skills, which involve establishing coordination between multiple limbs.

Besides, as it is well known that technology has been playing a considerable role in learning motor skills [9], the procedure of this learning is based mainly on direct verbal cooperation between the educator and students. The educator provides explanatory instructions which are an important tool for the transfer of this information, especially when the claim of synchronization is important for the performance of this motor skill [7]. Additionally, verbal cues that highlight important components of the skill can be very useful additions to a modeled performance [26].

In gymnastic skills, which are considered high in complexity (the number of components involved) but low in organization (the extent to which the components are interrelated) [13], the subject is obligated to perform with the “correct body position” especially in the supported phases of the skill. Additionally, in serial skills, the learner must be focused on doing the earlier parts more consistently, which simplifies the control needed on the subsequent phases [18, p. 269].

Especially in the vaults on the vaulting horse, the angular position of the body relative to the apparatus in the support phase on the horse (1st flight phase, the passing of the body from the vertical line of the horse, push-off phase on the horse) must be identified with the technical aspects of the skill in order to create the prerequisites for successful performance. In these cases, the instructions related to the synchronization, coordination of each limb (internal coordination) and external coordination of these limbs (position of these limbs in relation to space) play an important role in the technical performance of this skill.
Previous findings that referred to biomechanical studies with elite gymnasts revealed the correlation between kinematic variables and the judges’ score [22]. Another view of interest on motor control and learning which examined the procedure of learning various motor skills [24] and referred to the relationship between the speed during run-up phase and the final score of the vault [4].

Especially in the vaults on the vaulting horse, which consist of many different phases, the influence of one phase of the skill on the next is not well known, nor is the relationship between the body positions during the performance of the skill. So far there has been no conclusive examination of the relationship between the body limbs related to the angular positions of the body on which depends the differentiation of the frequency of the kinematic information feedback during acquisition that influences the learning and the retention of learning in these positions. This is also not very clear from the motor control and learning perspective, and therefore the purpose of this study was to examine: a) the relationship between various parameters of the support phase of horse vaulting and the ability to maintain the level of learning the skill and b) the existence of differences between groups and measurements. For this reason the purpose of this study was to examine which components are retained in the subjects’ memory after the post test of performing the handspring on the vaulting horse, as well as the amount of differentiation between those parameters during retention of learning in relation to the post test.

Materials and methods

Samples
Forty-five male novice undergraduate students at the Department of Physical Education and Sports Science, aged between 18-23 years old (20.60±1.29) without previous experience of vaulting horse skills participated in this study. The subjects were assigned randomly into three equal groups (N=15) and they pledged not to participate in this or any other skill beyond the time of the experimental sessions. The somatic characteristics of the subjects according to their group are presented in Table 1.

The three groups received a) 100% relative frequency of kinematic information (after each trial), b) 66% relative frequency of kinematic information (after each trial for the first 9 trials and summary feedback after 3 trials for the last trials), and c) 33% relative frequency of kinematic information (summary feedback after 3 trials) while they learned the handspring vault on the vaulting horse. Each participant performed the task in individual sessions.

Material and Procedure

The skill that was performed was the handspring on the vaulting horse. Practice of the skill and measurements of the experimental procedure were done in a gymnasium with a screen behind which the subjects were waiting for their participation in order to exclude the possibility of observing the trials of the other subjects [2]. The apparatus used for the performance of the skill consisted of a SPIETH Reuther board, a JANSEN-FRITZEN vaulting horse and mats for safe landing. A JVC GR-AX2 video camera (25 fr/sec) with optical axis perpendicular to the horse was used to record the trials that were in turn evaluated by official judges of artistic gymnastics.

Experimantal Procedure
In the regime of experimental design the subjects participated in five preparative educational days, including basic prerequisite components of the experimental skill, while the next day the initial (PRETEST) measurement was done in order to identify the initial level of the skill. The practice program of the experimental groups started one day after PRETEST and included 12 sessions where the subjects performed 18 trials in each session. One day after the end of practice the final (POSTTEST) measurements were done in order to identify the level of learning of the skill, while two days after the POSTTEST the 1st retention of learning was done, followed by the 2nd retention of learning three days later. During practice (12 sessions) the subjects were advised to follow a special regime for skill learning and for each trial of skill performance they received two forms with a list of 20 technical instructions (Table 2).

Throughout the practice sessions the instructions were verbal [29] and they determined the basic technical aspects of the experimental skill. Verbal feedback referred to two most serious errors relative to the correct performance of the subsequent trial during the practice of the experimental skill [11, 30]. Practice was programmed to be done day by day, distributed so because these kinds of practice “guide in better performance and learning but benefits of performance are greater from these of learning” [10], as well as to provide sufficient time for rest. Before each practice session the subjects had 10-12 minutes warm-up period in order to increase the arousal levels, which had a beneficial influence on the subsequent practice session [6], and to minimize the possible decrement of performance (warm up decrement) for the rest day before the practice day [12].

The dependent variables referred to the angular characteristics on a body position during performance of skill [3] and determined i) the angle of the first flight phase (°), ii) the vertical body position passing the vertical axis (°), iii) the push off angle from the horse (°) and iv) maximum height of hips during the 2nd flight phase (cm). These parameters were calculated by a special computer video analysis program BIOKIN [5]. The stick plots were determined with a deviation of ± 1 frame.

Statistical Analysis
The raw data were analyzed with the statistical program SPSS v. 13.0. The mean of three trials performed was the dependent variable and simultaneously the final score of each subject for every one of the four measurements in the assessment of the experimental skill. In order to show the existence of significant differences between groups, the MANOVA method of four parameters was used with the group and measurement factors. The level of significance for all analyses was set at p = 0.05.

Results

The MANOVA procedure with the four measurements of skill assessment as the dependent variables, and group (3 groups) and measurement (post, ret1 and ret2) as the independent factors revealed that the group did not have any significant effect on the overall model, either by itself (Wilks’ Lambda=0.919, F8,246=1.33, p=0.228), or through its interaction with the measurement (Wilks’ Lambda=0.883, F16,376=0.98, p=0.476). Contrarily, the measurement did have a significant effect (Wilks’ Lambda=0.480, F8,246=13.66, p<0.001).

As Figure 1 and Table 3 show, the significant effect of measurement is manifested only in one of the four measures, namely the vertical position (F2,126=62.6, p<0.001). The effect size (expressed here as eta squared) was high (49.8%).

The results of stepwise linear regression of the four parameters at different measurements are presented in Table 4. The values reported are the multiple correlation coefficients (R), the coefficients of determination (R2) in italics and the p-values in brackets.

In the post measurement, vertical position had a significant positive correlation with angle1 (the angle of the 1st flight phase). In the two retention measurements the correlation remained significant, although it became negative. Likewise, the significant correlation between vertical position and angle2 changed sign from negative in the post measurement to positive in the first retention measurement, while it was insignificant in the second retention measurement. A significant negative correlation between angle1 and angle2 was observed only in the first retention measurement. Finally, in the post and first retention measurements height was significantly negatively correlated only with angle2, while in the second retention measurement height was more negatively correlated with vertical position than with angle2. While the final skill (height) was positively correlated with the first skill (angle1), the variability of height could be better explained by the variability of its predecessors, angle2 and vertical position.

Discussion

Subjects performed and retained angle 1 similarly in all measurements. However, the lower body position (22°), relative to the ideal at this phase (Figure 2) appears to have had an effect in the next phase.

For example, according to laws of mechanics, a body position of 22° does not facilitate body rotation around the instantaneous axis of rotation on the vaulting horse apparatus during the supportive phase of hands [20]. In order to continue body rotation in the same direction, the subjects exhibiting the 22° position move their shoulders forward, leading to a transfer of their CG nearest to the rotation axis (contact point of hands on the horse). Overall, the decrement of the lever arm appears to facilitate the forward rotation of the body. The increment of the vertical position has a negative effect on the angle of repulsion from the horse as a result of the forward displacement of shoulders during the support phase on the horse (Table 3). The angle of repulsion is strictly related to the height of the 2nd flight phase that is characterized by absence of raising the centre of gravity of the body. According to the results in Table 4, vertical position of the body depended positively on angle 1 in post test and negatively in two retention measurements (Ret1 & Ret2). This means that an increase of the 1st flight phase (angle1) is followed by an increase in vertical body position, which reverses in Ret1 and Ret2.

The fact that angle 1 remains similar in three measurements (22°, 23.1° and 23.9° for the post test, retention 1 and retention 2, respectively) leads us to the conclusion that athletes may achieve ‘vertical position’ when performing vaults with the higher angle 1 as a result of the lever arm decrement [20]. On the other hand, the relationship between the angle and the vertical position was reversed in the next measurements (retention 1 and retention 2) and may be attributed to the fact that ‘vertical position’, which is achieved in ‘temporal delay manner’ and has a negative relationship with angle 1, is affected by other mechanical factors, such as run up velocity. Overall, the athlete’s multiple segment joints affect the ‘vertical position of the body’ which, in turn, may be attributed to different placement of hands (the angle between the hands and the torso). That means each subject must perform counterbalance movements with the rest of the body, e.g. blocking shoulders, moving the legs faster, etc. Further, according to Gibson [8], Turvey and Carello [23], the subject/athlete is more capable of producing a solution for each motor problem. Motor behavior is a complex blend of open and close loops and related with the duration of the skill. In case that motion is very fast it appears to be organized in advance and carried out without many modifications from sensory feedback [17].

Possibly, the experience from the previous day and the athletes’ ability to recall information from practice were not retained and therefore the relationship between the angle and vertical positions was reversed during the retention measurements. Vertical position exhibited a negative relationship with angle2 (Table 4), meaning that an increment in the vertical position results in the decrement of angle2 in post test. On the other hand, the above negative relationship was reversed in retention 1, which may be attributed to the different body position during the vertical position. On the contrary, height was not affected by angle1 or/and angle2 in all measurements. It is obvious that there is a negative relation between angle2 and height of 2nd flight phase. The results showed that difficulties arising with regard to correct body position related to the support phase of the horse at the instant when the body passes the vertical level. This situation reinforces the technical view of the coaches concerning the shoulder blocking phase during the push off phase on the horse. This aspect is supported by the fact that in skills with many degrees of freedom the requirements are increasing and with body’s momentum during 1st flight phase - at the instant when the hands are supported on the horse - they do not have the possibility to control this momentum. This results in the forward motion of the shoulders, with the angle position of the hands with respect to the body, meanwhile the forward rotation of the body continues to the vertical level. The differentiation of vertical position of the body on angle1 (from positive in the post test to negative in the two retention tests) may be due to the different position of the whole body. It is well known that two learners may have the same score in a motor skill independently of the number of errors or the magnitude of these errors. In this case the learner differentiates the relation between the parts of the body, which depicts a different position of the whole body (posture).

Conclusions

Kinematic information feedback (verbal instructions about the pattern of motion in the experimental skill) is an important variable of learning serial gymnastic skills (handspring in the vaulting horse). The inability to perform the earlier phase of this skill affects the subsequent phases and this is in accordance with Schmidt’s theory [18] as well the claim by Naylor and Briggs [13], which maintain that in gymnastic skills, which are considered high in complexity but low in organization, the subjects are obligated to perform with a “correct body position” especially in the supported phases of the skill. In the present study, “vertical position” is the most important phase because it represents the base for correct continuation of the skill at the repulsion phase of hands on the horse (angle2). For this reason learners must practice the skill with other helpful motor skills, e.g. special pre-exercises which reinforce the correct position of the body in the different phases of the vault; they also must use different methods in practice in order to enhance the whole performance of the skill. This situation agrees with the relevant literature which states that in motor skills with many degrees of freedom, segmentation involves sequencing the skill according to certain spatial and/or temporal criteria [27]. For this reason it is recommended to carry out a study with a group practicing with respect to performance errors and use special helpful exercises, in order to clarify the influence of these exercises on the correction of performance as well as the influence of learning and retention of learning on the experimental skill.

Although feedback has a positive effect on motor skills learning, more specialized forms such as kinematic information feedback can be more effective as the subjects’ information about their performance is based on technical instructions of the pattern of motion.

Practical Application

The information gained from this study further supports the notion that training based on technical instructions of the skill provides specialized information to subjects about errors in performance. This means that although feedback is considered a useful tool to improve performance of motor skills, subjects using the appropriate verbal kinematic information feedback correct subsequent trials according to the pattern of this skill. In addition, subjects informed by these technical instructions are capable, by themselves, to compare their own performance with the pattern of this skill.