ARAT MFM, D3 (%) 9HPT(sec) Jebsenm(sec)
  Total Scale D     J1 J2 J3 J4 J5 J6
ID 1
Trained Untrained 6
-16		 1
-4		 4.8
-4.8		 5.6
2.4		 ND T2
ND T4		 ND T2
ND T4		 ND T2
ND T4		 ND T2
ND T4		 ND T2
ND T4		 ND T2
ND T4
ID 2
Trained Untrained -2
-4		 -2
-2		 0
-9.5		 0.4
-17.0		 U T4
2.5		 U T2/T4
10.3		 U T2/T4
U T2/T4		 -1.8
-1.7		 U T4
-2.3		 U T2/T4
U T2/T4
ID 3
Trained Untrained -13
-21		 -2
-2		 0
-4.8		 -7.2
1.1		 -3.3
-0.7		 0.8
U T4		 U T4
U T4		 -1.7
-3.3		 5.4
1.3		 2.3
2.0
ID 4
Trained Untrained 0
-9		 0
0		 0
0		 -8.0
-4		 19.9
-4.9		 -21.6
U T4		 U T4
U T4		 -0.7
-2.6		 U T4
U T4		 U T4
-10.7
ID 5
Trained
Untrained 		-14
-8		 -1
-1		 4.8
0		 -4.8
-10		 -0.6
0.8		 -1.7
-4.0		 -7.5
-1.9		 -2.0
-4.4		 -0.4
-1.6		 -2.8
-1.1
ID 6
Trained Untrained -2
2		 0
0		 0
-4.8		 -4.1
14.4		 -0.6
-4.8		 -0.5
1.3		 9.5
U T4		 0.8
0.8		 0.6
-5.1		 U T4
U T4
Table 3: Effects of 24 weeks of training with dynamic arm support (N=6). Data are individual differences in scores between T2 (start of training) and T4 (primary endpoint) for the trained and untrained arms. ARAT: Action Research Arm Test; MFM: Motor Function Measure; 9HPT: Nine-Hole Peg Test; Jebsen: Jebsen Taylor Hand Function Test; J1: Turning cards; J2: Small objects; J3: Feeding; J4: Stacking; J5: Light objects; J6: Heavy objects; ND: Not determined at the indicated assessment(s); U: Unable to perform at the indicated assessment(s).