Effects due to Obstacles

All temporal and spatial dependent measures showed a significant effect due to object configuration and are summarized in Table 2.1. As predicted from previous work (Chapman & Goodale, 2008), the effects observed due to the positioning of non-target objects between the participant and the reach target demonstrate that the objects were functioning as obstacles, caused interference, and were being avoided. Configurations with objects close to the right (reaching) arm of the participant (2-B-In, 2-R-In, 1-R-In) caused the most interference, yielding longer overall movement times (longer in both the early and late phase of movement) and lower peak velocities. Configurations with objects placed further away on the right (1-R-Out, 2-L-In) caused some interference while configurations with objects only on the left, or no objects (1-L-In, 1-L-Out, None) caused the least interference as characterized by these temporal measures (see Table 2.1). Reaction time effects were largely due to the longest reaction times occurring on trials with no objects, an effect we attribute to the relative novelty of these trials.

Table 2.1 – Dependent measures showing an effect due to obstacle Configuration

Rxn Time (ms)

2 B-In 2 R-In 2-L-In 1 L-In 1 R-In 1 L-Out 1 R-Out None F

372.85 373.48 358.26 377.12 384.13 374.53 367.20 393.71 ** 2-L-In < 1-R-In,None;

Mvmt Time (ms)

2 B-In 2 R-In 2-L-In 1 L-In 1 R-In 1 L-Out 1 R-Out None F

669.25 663.00 648.24 638.09 668.19 638.17 647.91 641.23 ** 2-B-In, 2-R-In, 1-R-In > 1-L-In, 1-R-Out; 2-B-In, 1-R-In > 2-L-In, 1-L-In, 1-L-Out, None;

Time To PV (ms)

2 B-In 2 R-In 2-L-In 1 L-In 1 R-In 1 L-Out 1 R-Out None F

267.34 270.57 264.56 255.70 271.97 258.94 264.33 257.85 ** 2-R-In,1-R-In > 1-L-In,1-L-Out,1-R-Out,None; 2-B-In,2-L-In,1-R-Out > 1-L-In; 2-B-In > None;

Decel Time (ms)

2 B-In 2 R-In 2-L-In 1 L-In 1 R-In 1 L-Out 1 R-Out None F

401.91 392.43 383.68 382.39 396.23 379.22 383.58 383.38 ** 2-B-In > 1-L-Out; 1-R-In > 1-R-Out;

Peak V (mm/s)

2 B-In 2 R-In 2-L-In 1 L-In 1 R-In 1 L-Out 1 R-Out None F

1487.75 1498.59 1535.71 1579.85 1505.01 1594.21 1554.99 1587.73 ** 2-B-In,2-R-In,1-R-In < 1-L-In,1-L-Out,1-R-Out,None; 2-B-In < 2-L-In; 2-L-In < 1-L-Out;

X@100 (mm)

2 B-In 2 R-In 2-L-In 1 L-In 1 R-In 1 L-Out 1 R-Out None F

-7.29 -11.26 -1.80 4.38 -14.19 0.76 -6.76 -0.68 ** 1-R-In < 2-R-In < 2-B-In,1-R-Out < 2-L-In,None < 1-L-Out < 1-L-In;

X@250 (mm)

2 B-In 2 R-In 2-L-In 1 L-In 1 R-In 1 L-Out 1 R-Out None F

-12.00 -17.65 -2.26 7.17 -21.33 1.92 -9.16 -0.15 ** 1-R-In < 2-R-In < 2-B-In,1-R-Out < 2-L-In,None < 1-L-Out < 1-L-In; X@End NV (mm) 2 B-In 2 R-In 2-L-In 1 L-In 1 R-In 1 L-Out 1 R-Out None F

1.22 1.95 -0.88 -0.54 1.60 -1.06 -0.16 -0.74 * X@End V (mm) 2 B-In 2 R-In 2-L-In 1 L-In 1 R-In 1 L-Out 1 R-Out None F

2.59 3.67 3.28 3.42 3.42 3.25 3.30 3.19 NS

The ‗F‘ column shows an F-test of the main effect. Where significant, results from pairwise contrasts for each dependent measure are shown below the compared means. * & < or > = p<0.05, ** = p<0.005. NS = not significant.

Patterns of lateral deviation early (X@100) and mid-reach (X@250) for trials with two objects matched our previous work (Chapman & Goodale, 2008). The trajectories with the 2-R-In objects were shifted the most leftward and the trajectories with the 2-L-In were shifted most rightward. The trajectories with the 2-B-In fell in between (see Table 2.1 and Figure 2.2). The current experiment makes two important additions in

quantifying how we react to obstacles. First, when compared to two-object trials, the deviations on one object trials were larger. For example, the 1-R-In trials resulted in leftward deviations that were even larger than those observed on the 2-R-In trials, indicating that the second object in the two-object trials was, not surprisingly, having an effect. This pattern was observed for objects on both sides of space (see Table 2.1 and Figure 2.2). Second, when making reaches to a specific point in space (as opposed to the target strip used in previous work), the lateral deviation at the endpoint (X@End) was largely unaffected by the object configuration. We did observe, however, an interaction between Feedback and Configuration for the X@End measure F(5.27,94.91) = 3.50, p<0.01. Simple main effect follow-ups revealed that there was no significant effect of configuration on V trials and a small but significant effect of configuration on NV trials. Specifically, it appears that on NV trials there is an overshoot, whereby the finger ended slightly further to the right on trials where the reach was initially pushed to the left (e.g. 2-R-In) than it did on trials where the reach was initially pushed to the right (e.g. 2-L-In). Overshooting the target along the primary axis of movement on NV trials has been shown previously (e.g. Heath et al., 2004), and our results are likely an extension of these

Table 2.2 – Dependent measures showing an interaction between Feedback and Schedule

Mvmt Time (ms) Interaction* feedback x schedule Blocked Random Alternating F

No Vision 635.25 672.68 655.51 * Blk<Rnd With Vision 654.02 651.67 642.36 NS - Decel Time (ms) Interaction* feedback x schedule

Blocked Random Alternating F

No Vision 376.42 403.29 394.91 * none With Vision 387.97 383.46 381.26 NS - Peak V (mm/s) Interaction* feedback x schedule

Blocked Random Alternating F

No Vision 1607.11 1514.49 1542.81 * Blk>Rnd With Vision 1511.65 1528.91 1550.55 NS - X@250 (mm) Interaction* feedback x schedule

Blocked Random Alternating F

No Vision -9.56 -5.27 -6.21 * Blk<Rnd With Vision -6.87 -5.25 -7.08 NS - Std X@End (mm) Interaction* feedback x schedule

Blocked Random Alternating F

No Vision 10.55 9.56 8.85 * none With Vision 4.79 4.27 4.22 NS -

The strength of the interaction is indicated in the row with the measure name. F-tests results are shown comparing simple main effect means of each level of Feedback row across the three levels of Schedule. Results from pairwise contrasts are shown next to each significant F-test. * & < or > = p<0.05. NS = not significant.