
 
possible. It consists of two switching points; 
preliminary 5 s maintain both hands finger pinch 
force output at 10%MVF, then disappearing or 
appearing the feedback at 5 s (first switching point) 
and remained till 10 s, then second switching point 
at 10 s after the beginning of the measurement. The 
task continues 15 s and participants didn’t know the 
timing of switching point (Figure 1). 
 
 
Figure 1: Representative force outputs for two conditions. 
A: Force output of LUP&RKEEP condition. B: Force 
output of LUP&RDOWN condition. 1 indicates first 
switching point (after 5seconds) and 2 indicates second 
switching point (after 10seconds). Rectangle with broken 
line indicates analyzing area. Left hand force output 
exceeded 10.6%MVF on B, became overshoot error trial. 
 
Figure 2: Number of overshoot error. Graph above shows 
a number of left and right overshoot error separately. 
Arrows inside the circle denote UP side while arrows in 
the middle of them indicate the opposite sides required 
force production (UP, DOWN, KEEP and WAIT). Graph 
below shows the total number of overshoot error. 
2.3 Data Analysis 
We focused on second switching point (Figure 1). 
There are 4 different force producing types: increase 
force output from 0 to 10%MVF (UP), decrease 
force output from 10 to 0%MVF (DOWN), keep 
maintaining 10%MVF (KEEP) and waiting at 
0%MVF force output level (WAIT). Combination of 
them with both hands, there are 7 different 
conditions: LUP&RUP, LUP&RDOWN, 
RUP&LDOWN, LUP&RKEEP, RUP&LKEEP, 
LUP&RWAIT and RUP&LWAIT. Participants 
engaged three trials on each condition, arranged in 
random order. The error trial was defined as when 
UP side of the hands’ force output exceeded 
10.6%MVF before 2 s from second switching point. 
We calculated the error ratio as the number of error 
trials per total numbers of trial. 
3 RESULTS 
Error ratio of each condition was shown in Figure2. 
The number of error ratio was the fewest on 
UP&UP, total 25/198 (Left: 17/96 Right: 8/96), 
followed by KEEP&UP total 52/192 (Left: 28/96, 
Right: 24/96) and WAIT&UP total 64/191 (Left: 
40/96, Right: 24/95). Most mistakable condition was 
DOWN&UP, total 81/192 (Left: 50/96, Right: 
31/96). A chi-square test of independence was 
performed to examine the relationship between 
experimental tasks and overshoot error ratio. The 
relation between these variables was 
significant,Χ
3,   767
 42.3,   0.05. 
4 DISCUSSION 
When controlling one side of limb’s force output to 
certain target, overshoot error is influenced by the 
other side’s force producing type. It is easy to 
control target force output when both hand’s task are 
same force producing type. In contrast, it becomes 
difficult to control force output when anti-phase task 
is required on the other side. This may because the
 
attention to the difficulty and complexity of the 
other side’s limb movement and/or task disturbs the 
control of the force output.  
The overshoot error ratio in right hand was lower 
than that of left hand. Certainly, the participants 
were all right-handed. Dominant limb is specialized 
for dynamic, feed-forward controlled unimanual 
tasks (Sainburg, 2002). There was relatively lower