Statistical package for the social sciences v 13

Values are expressed as mean ± SD or percentage. Differences between study groups were analysed using ANOVA with Bonferroni post-hoc analysis, Student t or chi-square tests. Relationships between variables were evaluated by Pearson correlation and multiple linear regression or multiple logistic regression models.
Kaplan-Meier curves and log rank tests of both mortality and hospitalizations were performed after stratifying by analysis subgroups. On multivariate Cox regression analysis, variables were included if they were independently associated with both the outcome and the exposure (p < 0·05) or if they modified the risk ratio estimate for any of the remaining covariates (> 0·5% change). Survival models were always adjusted for age, sex, pack-years, body mass index, Charlson index and current treatment. As an additional analysis, Poisson regression with overdispersion correction by Pearson were used to assess the significance of the weighted rate ratios for hospitalization.
All effects were considered significant with a p value < 0·05. Statistical analyses were performed using the Statistical Package for the Social Sciences, v13.0 (SPSS Inc, Chicago, IL, USA) and SAS for Windows statistical software, v9.2 (SAS Institute, Inc., Carey, NC, USA).

Data analysis was performed using the Statistical Package for the Social Sciences (v13.0, SPSS Inc., Chicago, IL, USA). Descriptive statistics, Kolmogorov–Smirnov (normality of the distribution) and Levene’s (homogeneity of variance) tests were calculated for all experimental data before inferential testing. Changes in health-related physical fitness parameters were compared over the training period for players in the two experimental and control groups using two-factor (group x time) univariate analysis of variance (ANOVA). If the appropriate statistical significance was identified, then the Bonferroni post-hoc test was used to further distinguish the differences among groups. Cohen d effect sizes (ES) were also calculated to determine the magnitude of the group differences in health-related physical fitness. ES was classified as follows: <0.2 was defined as trivial; 0.2–0.6 was defined as small; 0.6–1.2 was defined as moderate; 1.2–2.0 was defined as large; >2.0 was defined as very large; and >4.0 was defined as extremely large [39 (link)]. The Kolmogorov-Smirnov tests showed that data were normally distributed, and no violation of homogeneity of variance was found using Levene’s test. The statistical significance was set at p<0.05.