Statistical Tests for Sports Science Practitioners: Identifying Performance Gains in Individual Athletes.

There is an ongoing surge of sports science professionals within sports organizations. However, when seeking to determine training-related adaptations, sports scientists have demonstrated continued reliance on group-style statistical analyses that are held to critical assumptions not achievable in smaller-sample team settings. There is justification that these team settings are better suited for replicated single-subject analyses, but there is a dearth of literature to guide sports science professionals seeking methods appropriate for their teams. In this report, we summarize 4 methods' ability to detect performance adaptations at the replicated single-subject level and provide our assessment for the ideal methods. These methods included the model statistic, smallest worthwhile change, coefficient of variation (CV), and standard error of measurement (SEM), which were discussed alongside step-by-step guides for how to conduct each test. To contextualize the methods' use in practice, real countermovement vertical jump (CMJ) test data were used from 4 (2 females and 2 males) athletes who complete 5 biweekly CMJ test sessions. Each athlete was competing in basketball at the NCAA Division 1 level. We concluded that the combined application of the model statistic and CV methods should be preferred when seeking to objectively detect meaningful training adaptations in individual athletes. This combined approach ensures that the differences between the tests are (a) not random and (b) reflect a worthwhile change. Ultimately, the use of simple and effective methods that are not restricted by group-based statistical assumptions can aid practitioners when conducting performance tests to determine athlete adaptations. [ABSTRACT FROM AUTHOR]