An exploratory study to investigate the effects of motor performance and maturation on talent selection in handball

Talent selection, as an important component of talent development, is often examined in the literature using a variable- or person-centered approach in order to better understand which criteria are used in the selection decision. The present study compares the two approaches to investigate the link between motor performance and nomination status in youth. This included both the selection decision at state level (talent selection I) and the nomination for the national team trials (talent selection II). A sample of 331 male ( M Age = 12.96 ± 0.19) and 348 female ( M Age = 12.92 ± 0.29) handball players completed six motor performance tests (20-m Sprint, Jump and Reach, Shuttle Run, Push-Up Test, Throwing Velocity, Dribbling Test) as part of the state squad selection process. To measure maturation, anthropometric values of the athletes were collected and the Mirwald equation was used ( Mirwald et al., 2002 ). To analyze the link between individual test results (independent variables) and squad nomination (dichotomous dependent variable), a logistic regression model ( Nagelkerke ′ s R 2 , Odds Ratio) was used, differentiated by gender and squad nomination level (variable-centered). In addition, a cluster analysis (Ward's method, k -means) was performed using the person-centered approach. The relationship between the identified clusters and the nomination of executives was examined (Odds Ratio). Logistic regression showed that throwing performance had a significant influence on talent selection I for both boys and girls ( OR Boys = 1.8, p < 0.05; OR Girls = 2.73, p < 0.05). For girls, a significant influence on talent selection II ( OR = 1.93, p < 0.05) was also found. Based on the cluster analysis, four clusters were identified for boys and five clusters for girls. Across genders, athletes in one cluster are significantly more likely to be nominated for the State-Squad ( OR Boys = 3.71, p < 0.05; OR Girls = 3.67, p < 0.05). This cluster is characterized by above-average performance in all tests. In the case of girls, this link can also be demonstrated in talent selection II ( OR = 2.66, p < 0.05). The higher effect sizes of the person-centered approach compared to the variable-centered approach argue in favor of a more holistic view of motor skills in sport in the form of a profile analysis for talent selection.