2.6. Data Analysis

A 2-way mixed ANOVA was performed on sleep variables and bedtime KSS and VAS fatigue scores with “Group” (actors vs. non-actors) as between-groups factor and “Condition” (BL vs. TR) as within-groups factor. The same ANOVA was conducted on performance measures, as well as on KSS and VAS fatigue scores collected before training and recall, with “Group” (actors vs. non-actors) as between-groups factor and “Recall phase” (immediate vs. delayed) as within-groups factor. In case of significance, η² was used as a measure of effect size and the Tuckey test for post-hoc analysis.

Following statistical guidelines to correct for multiple testing without running a too high risk of Type II Error (see, e.g., [37]), we applied to sleep measures analyses an adapted Bonferroni procedure: the conventional alpha value (p ≤ 0.05) was divided by four, i.e., by the number of relevant sleep “dimensions” addressed in our research (“sleep classical measures”, “sleep continuity”, “sleep stability”, “sleep organization”). Therefore, significance was set at p ≤ 0.0125.

The alpha level was maintained at p ≤ 0.05 for the analyses concerning performance measures, sleepiness, and fatigue levels.

All analyses were conducted using JAMOVI 1.6.23 [38].

The data were analyzed by O.D.R., who had not participated in data collection.