. 2020 Sep 30;3:127. doi: 10.1038/s41746-020-00334-y

Table 2.

The minimum number of at-home monitoring days required to differentiate the at-home gait speed between younger and older groups similar to full data.

Days	Median gait speed			95^th Percentile gait speed
Days	t-Original	t-Bootstrap^b	p Value	t-Original	t-Bootstrap^b	p Value
1		3.01 ± 0.67	0.048		3.35 ± 0.68	0.08^a
2		3.57 ± 0.51	0.146^a		3.73 ± 0.53	0.147^a
3		3.79 ± 0.42	0.225^a		3.96 ± 0.41	0.214^a
4		3.91 ± 0.33	0.259^a		4.06 ± 0.33	0.243^a
5		3.99 ± 0.26	0.292^a		4.15 ± 0.26	0.288^a
all	4.13			4.29

For both 95^th and median gait speed 1000 bootstraps were drawn for various number of days [1–5] and the t-values were compared to the original t-values. In order to obtain similar group distinguishability compared to full data set, at least 2 and 1 days of at-home monitoring data with 1000 steps were required for median and 95^th gait speed, respectively (p > 0.05). t-Original shows the t-statistic for the group difference using the full data, t-bootstrap indicates the mean t-statistic over the bootstrap samples.

^ap < 0.05.

^bMean ± sd.