**3. Results**

#### *3.1. Control-Pre and Control-Post Conditions*

There was no significant di fference in performance score between the control-pre (Median = 2.975) and control-post (Median = 3.075) conditions (Z = −1.430, *p* = 0.153). Similarly, there was no significant di fference in the coe fficient of variation of performance score between the control-pre and control-post conditions (Z = −1.382, *p* = 0.167). We assumed that there was no pronounced carry-over or fatigue effect that significantly a ffected performance over the course of the experiment.

#### *3.2. Free-Throw Accuracy*

There were no significant di fferences in free-throw performance score ( χ2(4) = 6.510, *p* = 0.089) or the coe fficient of variation of the performance score ( χ2(4) = 5.629, *p* = 0.131) between the conditions (control-pre, top, bottom or full). However, post hoc pairwise comparison showed that the free-throw performance scores of the top (Median = 3.1, Z = −2.357, *p* = 0.018) and full (Median = 3.15, Z = −2.112, *p* = 0.035) conditions were significantly larger than that of the control-pre condition (Median = 2.975), as shown in Tables 2 and 3.


**Table 2.** Descriptive statistics of the averaged and coe fficient of variation of the free-throw performance score.

\* significant difference (*p* < 0.05) compared to the control-pre condition by post hoc Wilcoxon signed-rank test.

**Table 3.** Probability values (*p*-value) of the average (upper right triangle) and coe fficient of variation (lower left triangle) of the free-throw performance score.


\* significant difference (*p* < 0.05) by post hoc Wilcoxon signed-rank test.

#### *3.3. Full-Body Joint Range of Motion (RoM)*

One-way ANOVA repeated measures showed that the variation in compression garments imposed significant e ffects on the ROM of head flexion (*p* = 0.014, partial η2 = 0.169), trunk lateral bending (*p* = 0.024, partial η2 = 0.152), left shoulder flexion (*p* = 0.041, partial η2 = 0.152), right shoulder rotation (*p* = 0.048, partial η2 = 0.128) and left knee flexion (*p* = 0.003, partial η2 = 0.212). Post hoc pairwise comparison showed that the top condition significantly reduced the head flexion (*p* = 0.037; d = 0.503; 1.346, 95% CI 0.376 to 2.315) and trunk lateral bending (*p* = 0.042; d = 0.487; 1.039, 95% CI 0.041 to 2.036) ROM compared with the control-pre condition (Table 4). Similarly, the full condition significantly reduced head flexion (*p* = 0.009; d = 0.650; 1.346, 95% CI 0.376 to 2.315) and trunk lateral bending (*p* = 0.028; d = 0.532; 1.446, 95% CI 0.173 to 2.718) ROM compared to the control-pre condition.

**Table 4.** Descriptive statistics and one-way ANOVA repeated measures outcome of the range of motion of head and trunk in di fferent compression garmen<sup>t</sup> conditions.


FL/EX: flexion/extension; \* significant difference (*p* < 0.05) using one-way ANOVA repeated measures; g Greenhouse–Geisser correction to adjust the lack of sphericity; a and A denote *p* < 0.05 and *p* < 0.0125 than the control-pre condition.

Compared to that of the bottom condition, both the top (*p* = 0.01; d = 0.642; 3.422, 95% CI 0.929 to 5.915) and full (*p* = 0.003; d = 0.778; 3.530, 95% CI 1.405 to 5.655) conditions significantly reduced the ROM of the left shoulder flexion, while the top condition had significantly larger right shoulder rotation compared with the control-pre (*p* = 0.013; d = 0.611; 38.316, 95% CI −8.98 to 67.65) and bottom (*p* = 0.041; d = 0.491; 23.028, 95% CI 1.08 to 44.976) conditions (Table 5). The control-pre condition had significantly larger left knee flexion ROM than the bottom (*p* = 0.026; d = 0.539; 2.605, 95% CI 0.345 to 4.864) and full (*p* = 0.002; d = 0.804; 2.908, 95% CI 1.214 to 4.602) conditions. Similarly, the top condition had a significantly larger left knee flexion ROM than the bottom (*p* = 0.044; d = 0.482; 2.047, 95% CI 0.059 to 4.035) and full (*p* = 0.018; d = 0.585; 2.351, 95% CI, 0.469 to 4.232) conditions (Table 6).

**Table 5.** Descriptive statistics and one-way ANOVA repeated measures outcome of the range of motion of the upper limb in different compression garmen<sup>t</sup> conditions.


FL/EX: flexion/extension; AB/AD: abduction/adduction; RA/UL: Radial/Ulnar deviation; \* significant difference (*p* < 0.05) using one-way ANOVA repeated measures; g Greenhouse–Geisser correction to adjust the lack of sphericity; a denotes *p* < 0.05 than the control-pre condition; b and B denote *p* < 0.05 and *p* < 0.0125 than the top condition; C denotes *p* < 0.0125 than the bottom condition.

**Table 6.** Descriptive statistics and one-way ANOVA repeated measures outcome of the range of motion of the lower limb in different compression garmen<sup>t</sup> conditions.


FL/EX: flexion/extension; AB/AD: abduction/adduction; EV/IV: eversion/inversion; PL/DO: plantarflexion/ dorsiflexion; \* significant difference (*p* < 0.05) using one-way ANOVA repeated measures; g Greenhouse–Geisser correction to adjust the lack of sphericity; a and A denote *p* < 0.05 and *p* < 0.0125 than the control-pre condition; b denotes *p* < 0.05 than the top condition.
