A cross-validation study of the TGMD-2: The case of an adolescent population

Objectives: This study proposes an extension of a widely used test evaluating fundamental movement skills proficiency to an adolescent population, with a specific emphasis on validity and reliability for this older age group.

Design: Cross-sectional observational study.

Methods: A total of 844 participants (n=456 male, 12.03±0.49) participated in this study. The 12 fundamental movement skills of the TGMD-2 were assessed. Inter-rater reliability was examined to ensure a minimum of 95% consistency between coders. Confirmatory factor analysis was undertaken with a one-factor model (all 12 skills) and two-factor model (6 locomotor skills and 6 object-control skills) as proposed by Ulrich et al. (2000). The model fit was examined using χ², TLI, CFI and RMSEA. Test-retest reliability was carried out with a subsample of 35 participants.

Results: The test-retest reliability reached Intraclass Correlation Coefficient of 0.78 (locomotor), 0.76 (object related) and 0.91 (gross motor skill proficiency). The confirmatory factor analysis did not display a good fit for either the one-factor or two-factor model due to a really low contribution of several skills. A reduction in the number of skills to just seven (run, gallop, hop, horizontal jump, bounce, kick and roll) revealed an overall good fit by TLI, CFI and RMSEA measures.

Conclusions: The proposed new model offers the possibility of longitudinal studies to track the maturation of fundamental movement skills across the child and adolescent spectrum, while also giving researchers a valid assessment to tool to evaluate adolescent fundamental movement skills proficiency level.