70). The intercorrelations between the quadrupedal performances namely were more inconsistent and generally lower (from r=0.28 up to r=0.56) in comparison to the inter-correlations between bipedal performances (from r=0.53 up to 0.71). The factor analysis extracted only one significant factor with the highest projections of the S30 and BS30 (Table 2). Table 2 Intercorrelations between tests of linear speed and results of factor analysis Discussion The 30 m sprinting from a static start timed by electronic timing gates is often used for assessing running speed, and the sprinting results of our study are within the range of previously reported data (Young et al., 2008; Chaouachi et al., 2009; Green et al., 2011).

More precisely, our subjects achieved results better by 7% than non-elite rugby union players, and worse by 2�C3% than elite rugby union players, professional basketball players and elite Australian Rules footballers (Green et al., 2011; Chaouachi et al., 2009; Young et al., 2008). However, previous studies have not reported data for other test performances that were used in the present study and therefore, those results cannot be compared. Studies conducted so far reported high values of the inter-subject-reliability and within-subjects-reliability-coefficients of the sprint measures over distances up to 30 m (Chaouachi et al., 2009; Meylan et al., 2009; Green et al., 2011). Therefore, the high values of reliability parameters of 30-m sprint test in our study are in concordance with previous findings.

To the best of our knowledge, this is the first investigation that systematically studied the reliability of ��non-forward�� running speed tests. Evident differences can be observed between the CV values of quadrupedal and bipedal locomotions. Lower CV values for bipedal performances (1.3 to 3.2) compared to quadrupedal performances (6.6 to 9.1) are mostly explainable by familiarity of the tests. In contrast to quadrupedalism, bipedalism is a regular locomotion of adult humans. Familiarity with the movement patterns is one of the crucial factors for achieving the high reliability of the test procedures (Sealey et al., 2010). Therefore, it is logical that the subjects were more familiarized to more common activities (i.e., bipedal), which consequently led to a lower CV (i.e., higher reliability) for bipedal tests.

If only observing the factor analysis results, and significance of the correlation, one could conclude that S30 can be used as a universal tool for testing linear speed of different forms of locomotion. However, S30 shares only 17 to 34% common variances (the values r2 converted into percentages) with other performances in this study, with GSK-3 the exception of 50% of the common variance shared with BS30. In general, a larger percent of non-shared variance (> 50%) indicates that two studied variables possess specific or at least relatively independent qualities (Huck, 2008).