TABLE 2.

Training volume variables with the effect size (ES): Motor imagery training^* vs. no-exercise control.

Volume variables	Number of studies	Effect size (CI)	Heterogeneity
			Q	df	p	I2
Training period (weeks)	22		58.86	24	<0.001	59.2%
3	6	1.52 (1.08–1.96)	2.52	5	=0.77	0.0%
4	7	1.16 (0.78–1.55)	6.00	6	=0.42	0.0%
8	6	1.07 (0.65–1.49)	23.19	5	<0.01	78.4%
12	3	1.95 (1.23–2.68)	1.44	2	=0.49	0.0%
Training frequency (per week)	22		58.86	24	<0.001	59.2%
3	6	1.06 (0.62–1.49)	11.40	5	<0.05	41.7%
5	16	1.39 (1.13–1.66)	25.73	15	<0.05	56.1%
Number of trials (per session)	22		58.86	24	<0.001	59.2%
20	3	1.18 (0.55–1.80)	9.80 2	2	=0.17	43.7%
30–40	11	1.48 (1.15–1.81)	19.82	10	<0.05	49.5%
50	6	1.05 (0.64–1.46)	17.41	5	<0.05	65.5%
100–120	2	1.41 (0.68–2.14)	0.02	1	=0.90	0.0%
Total number of trials (per study)	22		58.86	24	<0.001	59.2%
144–300	5	1.26 (0.77–1.75)	4.06	4	=0.40	1.4%
600	6	1.25 (0.83–1.68)	11.48	5	<0.05	56.4%
1,200	7	1.19 (0.80–1.57)	18.36	6	<0.05	67.2%
2,400–3,000	4	1.75 (1.16–2.33)	2.30	3	=0.51	0.0%

*Including only the 22 internal MIT studies in the analysis.