The Effect of Different Types of Feedback on Learning of Aerobic Gymnastics Elements

Abstract

The aim of our study was to discover the effect of different types of feedback in teaching/learning of difficulty elements of aerobic gymnastics. The study was designed as a one-group comparative interrupted time study. For this purpose, eight gymnasts competing in the national development category were selected (average age 9 ± 1.5 years; average sport age 4 ± 1.5 years). The design of the study included two intervention programs; each lasting 23 days interrupted by an 8-week break. In intervention program 1, the group practiced a straddle jump using self-modeling followed by verbal feedback, and in intervention program 2 the group practiced a split jump using expert modeling followed by verbal feedback. The experimental group practiced three times a week for a period of 30 min per usual training session (normally lasting 90 min). The results showed that the execution of the elements in both intervention programs improved at the 5% level of significance. In intervention program 1, we noted a significant improvement (p ≤ 0.05) in subphase approach and culmination. In intervention program 2, we noted a significant improvement (p ≤ 0.05) in culmination only. There was no significant difference between the changes of the final scores of the executed elements in intervention programs 1 and 2. Comparing the results of individual subphases, we noted no significant difference either.

1. Introduction

Aerobic gymnastics is a relatively new sport discipline in the family of gymnastics sports governed by the International Gymnastics Federation (FIG). This sport is characterized by the ability to perform continuous and high-intensity movements to music, originating from traditional aerobic movements. Competitors must demonstrate a high level of gymnastics and dance-based skills, difficulty elements, and acrobatic moves showing the sport’s variety and creativity [1,2]. The most successful gymnasts are those who manage to perform difficult routines with high accuracy and a proper technique [3].

In aerobic gymnastics, as well as in other sports, there is a demand to acquire new skills in the fastest and most effective way. Therefore, coaches and sport experts constantly try to find ways to ease and fasten the process of motor learning. To achieve a high level of trained skills, it is crucial to use feedback (FB) directed from the coach to the athlete. Thanks to a technological progress, the given FB is becoming more and more sophisticated [4]. Bilodeau et al. [5] stated that FB is the most important variable controlling learning and performance. This statement is also supported by Magill [6], who claims that skills are learned more quickly and performed better when athletes receive augmented FB during practice. It is not clear which type of FB is the most effective in practice of different skills. However, some researchers have identified different types of FB appropriate for various skills or stages of learning [7].

Nowadays, modeling is a common approach in learning new skills, especially in sport training [8]. Many authors agree that the most effective type of FB is model observation, which is based on imitation [9,10,11,12]. Observation enables an individual to determine the key features of the task, removing the need to create a cognitive representation of the action pattern through trial and error [13,14,15,16,17]. With recent technological progress, video has become a popular medium for obtaining qualitative information about sport performance during training and competitions [18,19]. The video facilitates the transfer of information from the coach to the athlete, especially if the coach cannot or is not able to provide a live demonstration [20]. Through video, learners can be provided with recordings of their own experiments (self-modeling) or video of another experienced athlete (expert modeling) [21]. Self-modeling allows the learner to get a better view of their own movements, while expert modeling tends to facilitate the acquisition of unknown movement patterns [22]. While observing expert models, a model is created in the learner’s memory, which is then used as a starting point for comparing and correcting their activities, which results in improved quality of performance, speeding up the learning process and mastering motor-control parameters [23].

However, many studies show that model observation without verbal feedback offers insufficient information to the learner [7,24,25,26]. By pointing out the positive or negative points of one’s technique, we make the learning process much more effective [27]. According to Hebert and Landin [25], the interaction between the visual and auditory stimuli is important in the learning process, allowing the athlete to focus on the relevant points of the practiced movement. These statements are also supported by McCullagh and Little [11], who claim that observation without verbal FB has a weak effect on the learning process. Also, the results of the review of Starzak et al. [28] indicated that combining verbal feedback with other types of feedback could be beneficial for learning different gymnastics skills.

It is also very important to mention that it is not just the instructions or feedback itself that determines their effectiveness, but also where the learner’s attention is directed [29]. Either, the learner can concentrate their attention internally (i.e., on the movement itself) or externally (i.e., on the object or the effect of the action) [30]. Although both of these methods provide increased enhancement of motor learning and performance [31,32], many previous studies proved larger effectiveness of external focus [33,34,35]. A majority of these studies, though, examined tasks that involved some kind of equipment, such as a golf club [36], a ball [37], or a baseball bat [38]. However, in sports where movement itself is a primary evaluation criteria (i.e., gymnastics, dance, synchronized swimming) some authors suggested that the performance of those types of skills may benefit more from an internal focus of attention [39,40]. This is also supported by other research that claimed different effects of attentional focus according to the characteristics of the performed task, evaluation method, and developmental stage [38,41,42,43].

Even though many studies [44,45,46,47,48,49,50] have proved that positive effects of model observation training are undoubtable, it is still unclear whether self-modeling or expert modeling is more effective in facilitating the learning process. As we wanted to contribute to the knowledge in this field, we decided to compare these two types of visual feedback followed by verbal feedback in the practice of aerobic gymnastics’ elements.

2. Materials and Methods

2.1. Participants

Eight gymnasts competing in the national development category were selected for the study (average age 9.0 ± 1.5 years; average sport age 4.0 ± 1.5 years). Selected gymnasts were attending the same sports club under the supervision of the same coach. All participants met the selection requirements for the study, which dictated that they had no previous experience learning selected aerobic gymnastics elements, they had been actively involved in aerobic gymnastics for minimum of 1 year, and they did not have any injury negatively influencing the experimental process. The procedures followed the ethical standards on human experimentation in compliance with the 1964 Declaration of Helsinki and its later amendments. The experiment was approved by the ethics committee of the Faculty of Physical Education and Sports at Comenius University in Bratislava (3/2020), and all participants provided informed assent and parental consent.

2.2. Study Design

The aim of our study was to compare the effect of self-modeling and expert modeling in the practice of difficulty elements of aerobic gymnastics. The study was designed as a one-group comparative interrupted time study. The design of the study included two intervention programs, each lasting 23 days and interrupted by an 8-week break. The group practiced three times a week for a period of 30 min per usual training session (normally lasting 90 min).

2.3. Intervention Programs

In intervention program 1 (IP1) gymnasts practiced a straddle jump (Figure 1) using self-modeling followed by verbal feedback. At the beginning of each training session (TS), the gymnasts were presented with a video of the last attempt from the previous TS. Subsequently, they performed the selected element in two series of three repetitions, while each attempt was followed by an individual verbal FB by the coach. After each series, the gymnasts were provided with a video of the best attempt of the series. After completing the second series and watching the video, they made another attempt with an effort to perform it avoiding mistakes. The video of the last trial was then presented to individual gymnasts at the beginning of the next TS. All videos were always presented once in standard speed and once in slow motion, while the coach always pointed out the key errors.

In intervention program 2 (IP2) gymnasts practiced a split jump (Figure 2) using expert modeling followed by verbal feedback. At the beginning of each TS, the gymnasts were presented with a video of the expert model. Subsequently, they performed the selected element in two series of three repetitions, while each attempt was followed by individual verbal FB by the coach. After each series, the gymnasts were again provided with a video of the expert model. After completing the second series and watching the video, they made another attempt with an effort to make it without mistakes. All videos were always presented once at standard speed and once in slow motion, while the coach always pointed out key factors of the element.

2.4. Materials

The practiced elements were recorded on a video camera before and after each intervention program. Before the initial measurements, the correct demonstration of the element as well as identical verbal instruction containing key factors to focus on in each phase of the element was presented. The video was recorded using a Canon SX 540 HS digital camera, with a resolution of 1920 × 1080 pixels and 60 fps. The camera was 3 m away from the tested person and moved 45° to the left. During each test, the gymnasts performed six attempts, of which the best three were selected for the expert technical analysis.

Self-modeling and expert modeling videos were presented using a tablet Samsung Galaxy Tab S7 (Samsung Electronics America, Ridgefield Park, NJ, USA) and the application Coach’s Eye, which offers an analysis of sports performance at normal speed as well as in slow motion.

2.5. Evaluation Process

Each gymnast performed six attempts (in both pretest and posttest) of which the best three were sent to four international aerobic gymnastics judges for expert technical analysis. The elements were divided into three phases and five subphases, each of which contained several evaluation criteria. In each subphase, judges assigned deductions for incorrect technical performance in accordance with the 2017–2020 Aerobic Gymnastics Code of Points [2]. In each subphase, a maximum of two points could be achieved. We divided the deductions based on the size of the errors as follows: small error—0.1 point; medium error—0.3 point; large error—0.5 point. The assigned deductions of each subphase were then deducted from the maximum score for the execution of the element—10 points. To ensure the consistency of the evaluation, all judges received precise instructions in advance describing how to assign the deductions in each subphase.

Analyzing the experts’ evaluations, we followed the international rules of aerobic gymnastics. In each subphase, we obtained four scores (one score from each judge), while we excluded the lowest and highest scores, and calculated the average from the two middle scores. The final score of individual gymnasts were obtained by calculating the average of the final scores from the three best attempts.

2.6. Statistical Analysis

Means and standard deviation of data were calculated to examine main indicators of descriptive statistics. The Shapiro–Wilk test for normality was performed on all variables. Data did not show a normal distribution of the traits; therefore, Wilcoxon’s signed-rank test was used to analyze the differences between pretests and posttests, as well as the effect of various types of feedback between intervention programs. The level of significance was set at p ≤ 0.05 and p ≤ 0.01. To interpret the practical significance of the research results, the effect size for nonparametric samples was calculated using formula: r = z/√N [51] (

Table 1. Interpretation of the effect size for nonparametric samples [51].

	Coefficient r
Effect Size	Positive	Negative
Small	0.10 ≤ r < 0.30	−0.10 ≤ r < −0.30
Medium	0.30 ≤ r < 0.50	−0.30 ≤ r < −0.50
Large	r ≥ 0.50	r ≥ −0.50

). Statistical and data analysis were performed using the statistical program IBM SPSS Statistics 23 (IBM, Armonk, NY, USA).

3. Results

3.1. Changes in the Performance of the Elements in IP1 and IP2

As shown in

Table 2. Changes in the performance of the elements in IP1 and IP2.

Subphase	Pretest		Posttest		Difference		p Value		Effect Size
	Mean	S.D.	Mean	S.D.	Mean	S.D.
Approach
IP1	1.87	0.08	1.93	0.05	0.06 *	−0.03	0.03		r = −0.53
IP2	1.86	0.06	1.92	0.05	0.05	−0.01	0.08		r = −0.45
Takeoff
IP1	1.67	0.18	1.77	0.09	0.10	−0.09	0.16		r = −0.35
IP2	1.61	0.06	1.67	0.11	0.06	0.06	0.18		r = −0.33
Culmination
IP1	1.10	0.49	1.46	0.36	0.36 *	−0.14	0.03		r = −0.55
IP2	1.09	0.36	1.40	0.33	0.31 *	−0.03	0.02		r = −0.60
Descent
IP1	1.53	0.14	1.63	0.17	0.10	0.03	0.09		r = −0.42
IP2	1.66	0.12	1.72	0.20	0.06	0.08	0.26		r = −0.28
Landing
IP1	1.51	0.18	1.49	0.23	−0.01	0.04	0.89		r = 0.04
IP2	1.49	0.29	1.61	0.23	0.12	−0.06	0.18		r = −0.34
Final score
IP1	7.68	1.08	8.28	0.90	0.60 *	−0.18	0.04		r = −0.53
IP2	7.72	0.89	8.33	0.93	0.61 *	0.04	0.02		r = −0.60

* for p ≤ 0.05.

, the group’s performance of the selected elements improved significantly in both intervention programs. Concerning the subphases of the element, in IP1 we noticed a significant change in the approach and culmination; however, in IP2 the significant change was noticed in culmination only. The biggest individual progress in the most important subphase—the culmination—is shown in Figure 3 (IP1) and Figure 4 (IP2).

3.2. Comparison of the Effect of Various Types of Feedback on the Performance of the Elements

In both intervention programs, there was a significant change in the execution of the performed elements. However, we noted no significant difference in the changes of either final score (Figure 5) or scores of the subphases (Figure 6) between the intervention programs.

Table 3. Differences expressed by the effect size.

Subphase	r	Effect Size
Approach	0.04	x
Takeoff	0.03	x
Culmination	0.07	x
Descent	0.14	small
Landing	−0.32	medium
Final score	0.11	small

shows the differences expressed by the effect size.

4. Discussion

The aim of our study was to discover the effect of various types of feedback in practice of aerobic gymnastics elements. The results indicated that the use of visual feedback together with verbal feedback is beneficial for learning new skills. However, we have not proven whether expert modeling or self-modeling is more effective, as in our study both methods revealed almost identical results.

We noticed several studies focusing on comparison of expert and self-modeling in the process of learning new skills in different sports. Firstly, there are studies claiming that expert modeling provides greater efficiency. For example, Zetou et al. [52] concluded that both self- and expert modeling had a positive effect on practice of new volleyball skills in young children. Moreover, they discovered that the improvement in the group of expert modeling was more significant. They explained that probably the expert model gave more accurate information or concentrated children’s attention more and motivated them to imitate and to strive more towards a better performance. This statement is also supported by Lirgg and Feltz [53], who claimed that subjects viewing a skilled model performed better and more efficiently than subjects who viewed unskilled models. Zetou et al. [54] conducted a similar study, repeatedly comparing the effect of expert and self-modeling in teaching volleyball skills to novices and came up with the same conclusion. On the other hand, some authors obtained results where self-modeling was more efficient than expert modeling. Kazakas et al. [55] compared the effect of self-modeling, expert modeling, and verbal feedback in the practice of badminton serves. All groups improved in performance, though the self-modeling group’s progress in the execution of the task was the most significant. Slightly different results were revealed in the study of Oñate et al. [56]. Their study focused on the right landing technique in basketball as a prevention of knee injuries. They compared self-modeling, expert modeling, and their combination, and the results showed that only the expert modeling group’s performance did not change significantly. Several authors [57,58,59] explained that self-modeling can facilitate the correction of technical faults and subsequent learning by influencing the development of a memory representation of the learned skill.

Some of the authors, though, agree that the most effective feedback method is a combination of expert and self-modeling. This important finding was conducted by Rohbanfard and Proteau [47], who stated that mixed observation of expert and novice models resulted in a better generalization of learning. These findings were also supported by Barzouka et al. [60], who compared these two methods and their combination in the practice of volleyball serves. Even though the execution of the skill improved in all experimental groups, the most significant change occurred in the group of combined feedback. Also, Robertson et al. [61] confirmed the same results, while they tested these methods in the process of learning new elements in artistic gymnastics, as well as Nishizawa and Kimura [62] in golf.

Beneficial effects of different video feedback methods have been conducted in the field of gymnastics as well. Baudry et al. [45] investigated whether video modeling can enhance gymnasts’ performance of the circle on a pommel horse. The results of the study revealed that the group who received expert and self-modeling and performance feedback improved their performance better than the control group. Robertson et al. [61] compared self-observation alone and self-observation combined with the viewing of a skilled model in the process of learning of two different gymnastics skills. Analysis of the results revealed that there were no differences between the intervention and early in acquisition, but later in acquisition, those skills practiced with the self-observation combined with viewing the skilled model were executed significantly better than those with only self-observation. Boyer et al. [21] analyzed the effect of a combination of the self-video feedback with a video of expert models in a group of gymnasts learning specific gymnastics skills. The results showed that all gymnasts demonstrated improved performance of the skills following exposure to the intervention. A different approach was taken by Amr-Dardari et al. [63], who compared the effects of different teaching/learning strategies (i.e., verbal feedback, video feedback with modeling, and video feedback with simulation) on performing basic vaulting skills on the vault table. They concluded that video feedback with model superposition had led to better learning improvements in the vault jump compared with simulation and verbal feedback methods.

All in all, we agree with the statement of Giannousi et al. [64], who declared that the most important factors in choosing the right feedback method are to take into consideration the age and experience of the learners. They assumed that adult novices would profit to a greater degree from observing themselves performing and younger novices would benefit more by observing an expert model, as they would try to imitate a proper pattern. Nevertheless, there is still a debate about the effect of different types of feedback. According to the available knowledge, we can assume that a key factor that significantly influences the performance of the practiced skill is the choice of the right method. Even though the results of our study did not reveal a difference in the effect of self- and expert modeling, we consider it important to find that both methods positively affected and facilitated the process of learning of aerobic gymnastics elements.

5. Conclusions

In conclusion, the results of our study provide evidence that both self-modeling and expert modeling followed by verbal feedback significantly improved the execution of the practiced elements in young aerobic gymnasts. Even though both methods displayed almost the same results, we assume that it is essential for coaches to include any of these methods during learning of new skills. Consequently, we conclude that our findings have important practical implications for athletes and coaches, who can benefit from the importance of using different types of feedback to improve athletic performance. Although this study extends the knowledge related to the effect of video feedback in the process of learning of new skills, we believe that some limitations should be acknowledged and future perspectives might be considered. Firstly, we realize that for future research, it is necessary to use a control group to assess the significance of the types of feedback used. Secondly, we advise extending the length of the experimental program to potentially increase its effectiveness. Lastly, to gather more information about the actual effect of selected types of feedback on performance and motor learning, it would be interesting to retest the subjects after the retention phase.