De Carvalho Barreta J.F.P.de S., et al.: GAZE BEHAVIOUR IN ELITE GYMNASTS WHEN Vol. 12 Issue 3: 287 - 297 GAZE BEHAVIOUR IN ELITE GYMNASTS WHEN PERFORMING MINI-TRAMPOLINE AND MINITRAMPOLINE WITH VAULTING TABLE - A PILOT STUDY Joana Filipa Pereira de Sousa de Carvalho Barreto1, Filipe Luís Martins Casanova2, César José Duarte Peixoto3 13 Faculty of Human Kinetics, University of Lisbon, Portugal 2 Faculty of Sport, University of Porto, Portugal _Case study Abstract Visual system provides information from the environment, leading gymnasts to improve performance. The question of what sources of visual information from the environment contribute to performance, remains unclear. This study aims to analyse visual behaviour, as areas of interest, fixated by Teamgym elite gymnasts during the performance of techniques on mini-trampoline and mini-trampoline with vaulting table. We hypothesised that: a) gymnasts would fixate areas of interest in the environment to visually perceive relevant information, b) fixations on the area of interest "landing mat" would occur during the last part of flight phase and landing. Three teamgym elite gymnasts performed three tasks on mini-trampoline and one task on mini-trampoline with vaulting table. The variables were: fixation duration (FD), areas of interest (AOIs), ratio between total fixation duration and total task duration (TFD/TD) and ratio between total fixation duration per AOI and total fixation duration (TFDA/TFD). Results showed that TFD/TD increased with the decreased of complexity degree for all tasks. Minitrampoline was the most fixated AOI (except for straight barani out on mini-trampoline) while wall was the less fixated. The run-up zone was the most fixated zone. For the task on minitrampoline with vaulting table, participants reduced the time spent fixating run-up zone and increased time spent fixating mini-trampoline and vaulting table. Landing mat was the only AOI that was fixated during the flight phase. Results suggest that gymnasts may adapt their visual strategy to the degree of complexity of the task, as reflected in the results of TFD/TD and TFDA/TFD. ... . Keywords: gymnastics, visual perception, eye-tracker. INTRODUCTION Vision provides information from environment in relation to the person, being part of a cycle where the perception of affordances guides behaviour and behavior discovers new affordances (Gibson, 1986). One of the contexts where this relation can be observed is in sports and, more specifically, gymnastics. The challenge is to analyze visual behavior of gymnasts, in an ecological environment, similar to practice or competition contexts. Vision is a dominant perceptual system assisting performance (Bardy & Laurent, 1998) on synchronize trampolining (Heinen & Czogalla, 2016), vaulting table (Heinen, Vinken, Jeraj, & Velentzas, 2013) and landing (Davlin, Sands, & Shultz, 2001; Luis & Tremblay, 2008). Also, the positions Science of Gymnastics Journal 287 Science of Gymnastics Journal De Carvalho Barreta J.F.P.de S., et al.: GAZE BEHAVIOUR IN ELITE GYMNASTS WHEN Vol. 12 Issue 3: 288 - 297 of springboard and vaulting table influence handspring kinematics (Heinen, Jeraj, Thoeren, & Vinken, 2011), suggesting that gymnasts visually perceive the positions of apparatus to adapt their movement. Such results were also confirmed for Yurchenko technique on vaulting table (Heinen, Brinker, Mack, & Hennig, 2017). Gymnasts perform better landings from back somersaults with full vision during performance, comparing to when vision is manipulated (Davlin et al., 2001). In synchronized trampolining, visual information from the other gymnast's arms seems to be a dominant cue to achieve synchronization (Heinen & Czogalla, 2016). These studies illustrate the importance of visual system to perform in diverse situations in gymnastics sport. Additionally to studies that revealed the importance of visual system in performing gymnastics, other studies investigated experts versus non-experts. Differences were found regarding the duration and location of fixations (von Lassberg, Beykirch, Campos, & Krug, 2012), (Mann, Williams, Ward, & Janelle, 2007). Experts reveal more significant correlations between gaze behaviour and kinematics of somersault, comparing to novices (Heinen, 2011). Expert gymnasts use visual information to regulate kinematic variables in every phases of movement, as: contact time with apparatus, flight duration, landing deviation and moment of inertia. An interaction between eye, head, spine and joint movements was shown, when gymnasts perform multiple twisting somersaults, pointing to a functional relation that facilitates orientation in space (von Laßberg, Beykirch, Mohler, & Bulthoff, 2014). Scientific evidences suggest a relationship between visual perception and movement adaptation in gymnastics. However, methodologies were conducted in a non-ecological way, which means that, it differs from practice/competition environments, namely in: similarity of the tasks performed, contextual constraints and in the use of indirect instruments to analyze visual behaviour. Some studies have manipulated visual behaviour by asking the gymnasts to fixate their gaze into a specific area (Heinen, Jeraj, Vinken, & Velentzas, 2012) or have occluded visual (Heinen, Koschnick, Schmidt-MaaB, & Vinken, 2014), (Luis and Tremblay, 2008) and auditory (Heinen et al., 2014) cues from the environment. Other studies measured gaze behaviour indirectly (not using an eye-tracker) and the tasks chosen were performed in isolation and not in a sequence as it is typical in training or competition contexts (Luis and Tremblay, 2008) (Heinen, Walter, Hennig, & Jeraj, 2018), (Davlin et al., 2001), (Bardy & Laurent, 1998), (Sato, Torii, Sasaki, & Heinen, 2017). Although this studies contribute to scientific knowledge in this domain, by evidencing a relationship between gaze behaviour and movement kinematics, analyze visual perception in real environment with direct instruments (ecological perspective), will provide more reliable details on visual patterns and on how gymnasts visually adapt to different contextual constraints. The aim of this pilot study is to analyse visual behaviour, as areas of interest fixated by Teamgym elite gymnasts during the performance of techniques on minitrampoline and mini-trampoline with vaulting table. We hypothesised that: a) gymnasts would fixate areas of interest in the environment to visually perceive relevant information, b) fixations on the area of interest "landing mat" would occur during the last part of flight phase and landing. METHODS Three Teamgym elite gymnasts (male, senior elite category, mean age = 21.6 years, SD = 5.4 years) took part in this study. All gymnasts had participated in Teamgym European Championships for the last four years. They reported normal or corrected to normal vision, and signed the informed Science of Gymnastics Journal 288 Science of Gymnastics Journal De Carvalho Barreta J.F.P.de S., et al.: GAZE BEHAVIOUR IN ELITE GYMNASTS WHEN Vol. 12 Issue 3: 289 - 297 consent form. They were able to perform the study tasks autonomously and were free from injuries. The study followed the guidelines of the Declaration of Helsinki and had ethical approval from Ethics Committee of Faculty of Human Kinetics (6/2018). The tasks were performed in i) minitrampoline and ii) mini-trampoline with vaulting table. Three tasks on minitrampoline and one task on mini-trampoline with vaulting table were analysed. The tasks included a 25 meters run-up towards the apparatus, the performance of the technique and landing. The techniques on minitrampoline were: straight barani (SB), tucked barani out (TBO) and straight barani out (SBO). On mini-trampoline with vaulting table, participants performed handspring straight barani out (HSBO). The tasks were chosen to represent characteristics of Teamgym discipline: 1) the two apparatus are performed in the same competition program (mini-trampoline). This program has a maximal duration of 2'45'' and the team performs 3 rounds with 6 gymnasts in each round. At least one round is performed with vaulting table. This implies that, in a short period of time, some gymnasts from the team perform in the two apparatus; 2) tasks analysed are similar in regarding movement: the last somersault and the half twist of SBO and HSBO are similar to SB; 3) tasks on mini-trampoline are used as a pedagogical progression to learn other tasks on mini-trampoline and mini-trampoline with vaulting table (example: SB is used as a pedagogical progression to learn the last phase of SBO and HSBO). Additionally, this tasks have body rotations in transversal and longitudinal axis, different body positions (tucked and straight) and an increase in degree of complexity, which makes this analysis richer. Complexity degree (difficulty score in Code of Points, CoP) is determined by degrees of rotation in longitudinal and transversal body axis and body position (Sjostrand, Lemmetty, Hughes, Gryga, & Jonsdottir, 2019) (Table 1). The Gymnasium at Faculty of Human Kinetics was prepared with apparatus according to Teamgym Directives from European Union of Gymnastics (UEG) (Sjostrand et al., 2019) (Figure 1). Tobii Pro Glasses 2 ® was worn by participants during the tasks. The system is a binocular eye tracker that records the point-of-gaze onto a video image of the scene, measuring the relative position of the pupil and corneal reflection. The recording process was undertaken using a Tobii Glasses Controller Software, running on a Dell Venue 11 Pro 7130, Windows 8/8.1 Pro tablet at a rate of 50 Hz. The image is transferred to a computer and analyzed by running the Tobii Glasses Analysis Software: Tobii Pro Lab. Being a direct method to study gaze behaviour, without manipulation of the visual stimulus (e.g. spatial occlusion), Tobii Pro Glasses 2 provided a higher level of ecology and revealed the expertise effects on gaze behaviour and decision-making (Kredel, Vater, Klostermann, & Hossner, 2017). Participants used a vest with a pocket placed on their back with the record unit. The vest was used under a fit t-shirt, to minimize the displacement of the record unit. Participants did a twenty-minute warm-up including general aerobic exercises of low intensity, specific exercises related to techniques of the study and some trials on mini-trampoline and minitrampoline with vaulting table. The warm-up protocol was similar to warm-up in training sessions. The eye-tracker was fitted onto de participant's head to perform three familiarization trials. After the adjustments, calibration was made by asking each participant to stand still and to fixate a target on the centre of the calibration card, at a distance of approximately 1.25 meters during 5 seconds. After calibration, to ensure the best quality of data, we asked the subjects to look to five different points in Science of Gymnastics Journal 289 Science of Gymnastics Journal De Carvalho Barreta J.F.P.de S., et al.: GAZE BEHAVIOUR IN ELITE GYMNASTS WHEN Vol. 12 Issue 3: 290 - 297 the environment. The five points were part of AOIs and were at various distances, heights and widths of the subject. Besides this procedure is recommended, it gives us an idea if the calibration is good or if there is a necessity to repeat it. Participants performed three trials of each technique, starting on minitrampoline. They were encouraged to perform tasks as in a competition. Intervals between twenty to forty seconds between trials allowed to verify the eye-tracker and to recalibrate when necessary. The gaze data were analysed frame by frame (sampling rate of 50Hz). Three gaze behaviour variables were considered for this study: number of fixations (NF), fixation duration (FD) and AOIs. A fixation was considered when gaze remained stationary for at least 99.99 milliseconds (Vickers, 1992), with a tolerance of 0.5° (Williams, Davids, & Williams, 2005). An Area of Interest/visual reference point is an area/object from the environment (for example, the minitrampoline) that is visually relevant for the participant when performing these tasks. Visual behaviour was analysed to calculate the percentage of viewing time, dividing the total fixations duration by task duration (ratio TFD/TD) and dividing the total fixations duration per AOI by total fixations duration (ratio TFDA/TFD). In this study, seven AOI were defined previously to analysis, based on objects that the participants need to physically contact to perform the tasks. Also, AOIs where the subject possibly would search for visual information were included (example: front and lateral walls). Lastly, "undefined" was considered an AOI in case any participant fixate an area not considered in the others. The seven AOIs considered were: a) 1st part of run-up, b) 2nd part of run-up, c) hurdle, d) mini-trampoline, e) vaulting table, f) landing mat, g) front and lateral walls and h) undefined. 1st and 2nd parts of run-up were 10 meters long and 2 meters large each, while hurdle which is the zone were gymnasts did the last foot contact with floor, were five meters long and 2 meters large. Landing mat was 4 meters large and 7 meters length. Wall was considered as the front and lateral walls and undefined was categorized when the fixation was not in any of the mentioned AOIs (Figure 1). Since this is a pilot study with an exploratory approach, with a small sample size and few repetitions, it was statistically underpowered, and only descriptive statistics were used (Field, 2018). Table 1 Complexity degree/Difficulty value for tasks analysed according to 2017-2021 Teamgym Code of Points - Juniors and Seniors. Task Straight barani Tucked barani out Straight barani Handspring straight _(SB)_(TBO)_out (SBO)_barani out (HSBO) Complexity degree/ 0.30 0.70 0.90 1.10 Difficulty value_ Science of Gymnastics Journal 290 Science of Gymnastics Journal De Carvalho Barreta J.F.P.de S., et al.: GAZE BEHAVIOUR IN ELITE GYMNASTS WHEN Vol. 12 Issue 3: 291 - 297 • 1 im ^m t T« !' 1 ' f) 9) lie] b) a} ft)