RELIABILITY AND VALIDITY OF A TEST BATTERY IN A SCHOLAR POPULATION

Many studies have examined the development of the motor skills during the evolutive age. So far, conditional tests have been mainly considered (Babin et al., 2001; Cintas et al. 2003), while there is a lack of information about coordinative capabilities tests. The aim of this study was to assess the reliability and validity of a battery of motor tests commonly used to evaluate children and adolescent people. 478 subjects aged 11 to 14 yrs (M:236; F: 242) performed six motor skills test (Carbonaro et al., 1988): dynamic balance test on a square beam (DBT), 10-m dash sprint (DS), standing long jump (SLJ), throwing of the basketball ball to assess explosive strength of upper limbs (SBT), sit and reach flexibility test (SRC), handgrip (HG). Body height (BH) and weight (BW) were also measured. The reliability was studied with the test-retest method, calculating the Pearson’s r correlation coefficient between trials of the same test, and the Cronbach’s Alpha coefficient. To assess the structural validity, hierarchical cluster analysis and a factor analysis (with MCA and VARIMAX rotation) were performed with the SPSS Software. Four tests (DS, SLJ, SRC, SBT) showed a very good test-retest correlation (r>0.90). Good correlation was observed for the HG (r=0.83) and the DBT (r=0.76). Separate analysis were conducted on subsamples divided by gender and age (11-12 yrs and 13-14 yrs), obtaining results similar, for both the test-retest correlation and Cronbach’s Alpha, to those observed in the aggregate sample. Hierarchical cluster analysis showed a first group of tests including HG, SBT together with BH and BW. The SLJ and DS tests were highly related each other and in a weaker way with the previous variables. Conversely, independence between the SRC and DBT and the other tests were observed. Factor analysis revealed three factors, explaining about 76% of the total variability. In the first factor (36% of var), anthropometric data, HG and DBT are highly saturated. In the second factor (27% var), the SLJ and DS are the most saturated variables. 13% of variability is explained by the third factor, correlated mostly to DBT and HG. The reliability of tests, assessed on a scholar population, showed good to very good levels. The structural validity analyses showed that the upper limb strength test is strictly related with BH and BW and then with the physical development of the boys. Strength tests involving a body displacement are instead less related to anthropometric characteristics. Coordinative tests as SRC and DBT are independent from both the strength tests and the physical development.