Table 3.

Comparison of mean and median rally length between women and men (reference group).

	Women		Men		Difference
	M (SD)	Mdn [IQI]	M (SD)	Mdn [IQI]	M (p, r)	Mdn [95% CI]
Rally length (s) with pseudo-rallies
Main event: ball in/out (n = 725 vs. 893)	7.25 (6.20)	4.97 [3.87, 9.57]	5.59 (4.67)	4.28 [3.12, 7.52]	1.66 (< 0.001, 0.16)	0.69 [0.42, 0.95]
Competitive event: fault (n = 37 vs. 59)	11.50 (9.00)	8.83 [5.37, 13.67]	8.25 (6.95)	5.03 [3.88, 10.08]	3.25 (0.049, 0.20)	3.80 [0.69, 6.91]
Pooled events: ball in/out + fault (n = 762 vs. 952)	7.45 (6.42)	5.03 [3.87, 9.83]	5.75 (4.88)	4.32 [3.20, 7.76]	1.70 (< 0.001, 0.16)	0.71 [0.43, 1.00]
Rally length (s) without pseudo-rallies
Main event: ball in/out (n = 614 vs. 714)	8.38 (6.07)	5.63 [4.33, 10.83]	6.79 (4.47)	4.76 [3.92, 8.44]	1.59 (< 0.001, 0.17)	0.87 [0.22, 1.52]
Competitive event: fault (n = 37 vs. 59)	11.50 (9.00)	8.83 [5.37, 13.67]	8.25 (6.95)	5.03 [3.88, 10.08]	3.25 (0.049, 0.20)	3.80 [0.69, 6.91]
Pooled events: ball in/out + fault (n = 651 vs. 773)	8.56 (6.31)	5.97 [4.33, 11.20]	6.90 (4.72)	4.76 [3.92, 8.48]	1.66 (< 0.001, 0.17)	1.21 [0.42, 2.00]

s = seconds; n = number of rallies (women vs. men); M = mean; SD = standard deviation; Mdn = median; IQI = interquartile interval; p = p-value of the Mann-Whitney U test; r = effect size measure r; CI = confidence interval of the difference between two medians obtained from bootstrapped quantile regression