INFORMACIÓN GENERAL Título del conjunto de datos: Dataset used for the study entitled: Segmental External Load in Linear Running in Elite Futsal Players: A Multifactorial and Individual Variability Analysis using Linear Mixed Models Fecha de recopilación de datos (fecha o rango de fechas): 2025-04-21 - 2025-01-30 Fecha de depósito: 2025-07-03 Idioma: español-inglés INFORMACIÓN DEL AUTOR Nombre: DIEGO HERNÁN VILLAREJO GARCÍA Institución: Universidad de Murcia Correo electrónico: dvillarejo@um.es ORCID: https://orcid.org/0000-0001-6149-9253 INFORMACIÓN METODOLÓGICA Participants: The study sample comprised 8 elite-level female futsal players (6 outfield players, 2 goalkeepers) (age: 29.9 ± 5.1 years, height: 164.96 ± 4.22 cm, body mass: 60.31 ± 4.56 kg) who participated in the Primera RFEF Futsal Femenina (Spanish first division) during the 2020/2021 season. All players were professional athletes with at least five years of experience in elite futsal and presented no musculoskeletal injuries or physical limitations that could have affected their performance during the study. The study protocol was approved by the Ethics Committee of the University of Murcia (approval number: 3180/2020). Equipment: For external workload assessment during the incremental running treadmill test, six WIMU PROâ„¢ inertial devices (Realtrack Systems, Almería, Spain) were utilized. These devices are equipped with triaxial accelerometers capable of detecting and measuring movement through a micro-electromechanical system. In the present research, the accelerometers' sampling frequency was set at 100 Hz. Data acquisition and initial processing were performed using the SPROâ„¢ software. A motorized treadmill (with a minimum capacity of 20 km/h and electronic speed control) was also used for the testing. To ensure proper sensor attachment, a back harness, elastic leg straps, and industrial-grade velcro were employed. Statistical Analysis: The PlayerLoad variable was subjected to a natural logarithm transformation prior to the analyses. This transformation was performed primarily to address the positive skewness in the distribution of external load and to improve the normality and homo-scedasticity of the model residuals [15]. To manage the high collinearity between speed and training zone, two separate Linear Mixed Model (LMM) analyses were employed. Both models included ID_jugadora as a random intercept to model individual variability among players. Additionally, Model 1 included Speed as a random slope to account for individual variability in the response to speed. Model 1 investigated the effects of Lo-cation, Speed, and their two-way interaction (Location*Speed). Model 2 investigated the effects of Location, Training Zones, and their two-way interaction (Location*Training Zone). Prior to LMM execution, statistical assumptions were verified for each model. The normality of the model's residuals was evaluated using the Kolmogorov-Smirnov test, complemented by visual inspection of Q-Q plots. For Model 1, the K-S test yielded p<.001; however, visual inspection of the Q-Q plot indicated a close approximation to normality, particularly in the central distribution, suggesting that the transformation substantially mitigated non-normality for practical inference given the large sample size. For Model 2, the K-S test yielded p=0.093; the Q-Q plot showed an excellent ap-proximation to normality. The homoscedasticity of residuals and the linearity of the relationship between predictor variables and the response variable were visually in-spected using scatter plots of residuals versus predicted values, with no significant de-viations observed in either model. Statistical significance for all fixed effects was set at p<.05. For a more complete interpretation of the practical magnitude of the effects, effect sizes were calculated using Partial Eta Squared (ηp2). These values were interpreted, where 0.01 is considered a small effect, 0.06 a medium effect, and 0.14 a large effect [34]. For a detailed interpretation of significant interactions, conceptual simple effects anal-yses were performed based on parameter estimates and confidence intervals. All sta-tistical analyses were carried out using Jamovi, version 2.6 [20] and its modules, along with supplementary computations in R Studio, version 4.5.0 [21] to derive the effect sizes (ηp2). Procedure: The research was conducted over three weeks, with one session per week. In the first week, participants were informed about the study's objectives and protocol. In the second week, they familiarized themselves with the treadmill test procedure and the experimental equipment (inertial sensors) by conducting a trial session. The incremental treadmill test was performed in the third week. Prior to the placement of the inertial devices, they were manually calibrated according to the manufacturer's recommendations and synchronized. This process is fundamental for eliminating 3D accelerometer error sources (offset error, scaling error, non-orthogonal error, and random error). The devices were affixed to the following anatomical locations following a validated scheme: (i) upper back, on the interscapular line; (ii) lumbar region, at L3 near the center of mass; (iii) knee, 3 cm above the kneecap; and (iv) ankle, 3 cm above the lateral malleolus. At both the knee and ankle, the devices were placed on the outside of the right leg on all athletes, while an anatomically designed harness was used for the upper back, and the rest were secured with a specifically designed elastic band. Before the test, participants performed a standardized 10-minute warm-up that included: 5 minutes of continuous running at 7 km/h, 3 minutes of joint mobility exercises, and 2 minutes of progressions (gentle accelerations on the treadmill). This procedure was monitored in real-time via WIMU PROâ„¢ devices that transmitted data wirelessly to the SPROâ„¢ software, ensuring the correct functioning of the sensors. The main protocol was an adaptation of the VAM (Maximal Aerobic Speed) test, designed for continuous and controlled analysis. The test began at a speed of 8.0 km/h, with increments of +0.2 km/h every 12 seconds. The test concluded upon the athlete's voluntary exhaustion or observation of a technical criterion (e.g., unstable stride frequency, evident fatigue). The treadmill surface was flat (1% incline). Upon completion of the main test, participants performed 5 minutes of recovery intensity running. To minimize interference from uncontrolled variables, all participants were instructed to maintain their habitual lifestyle and normal dietary intake before and during the study. Furthermore, tests were always performed at the same time of day (i.e., 9:00 AM) to avoid possible effects of circadian rhythms, and it was ensured that they had not performed high-intensity physical activity in the 48 hours prior to the tests. Environmental control was kept constant (temperature 22-24 ºC, humidity < 50 %). All participants used standard training footwear to homogenize conditions. The integrity of the sensors was verified at all times by two researchers. ARCHIVOS Nombre del archivo: descriptivo_modelo1.pdf; descriptivo_modelo2.pdf; MML_Modelo1.omv; MML_Modelo1.omv PALABRAS CLAVE External Load Monitoring, Female Athletes, Futsal, Inertial Measurement Units, Segmental PlayerLoad INFORMACIÓN DE PATROCINIO E IDENTIFICADORES DE SUBVENCIONES Sin financiación externa a la Universidad PUBLICACIONES RELACIONADAS Segmental External Load in Linear Running in Elite Futsal Players: A Multifactorial and Individual Variability Analysis using Linear Mixed Models LICENCIAS Y PRIVACIDAD Licencias:Creative Commons