PsychologicalResearch PsychologischeForschung

Psychol Res (1993) 55:291-298

© Springer-Verlag 1993

The role of impulse variability in manual-aiming asymmetries Richard G. Carson 1, Digby Elliott 2, David Goodman 1, Linda Thyer 1, Romeo Chua 1, and Eric A. Roy 3 School of Kinesiology, Simon Fraser University, Burnabyl BC, Canada 2 Department of Physical Education, McMaster University, Hamilton, Ontario, Canada 3 Department of Kinesiology, University of Waterloo, Kitchener, Ontario, Canada Received April 15, 1992/Accepted April 4, I993

Summary. Two experiments are reported in which we examined the hypothesis that the advantage of the right hand in target aiming arises from differences in impulse variability. Subjects made aiming movements with the left and fight hands. The force requirements of the movements were manipulated through the addition of mass to the limb (Experiments 1 and 2) and through control of movement amplitude (Experiment 1). Although the addition of mass diminished performance (i. e., it increased movement times in Experiment 1 and increased error in Experiment 2), the two hands were not differently affected by the manipulation of required force. In spite of the fact that the fight hand exhibited enhanced performance (i. e., lower movement times in Experiment 1 and greater accuracy in Experiment 2), these advantages were not reflected in kinematic measures of impulse variability.

Introduction In continuing attempts to account for the fight hand's advantage in manual aiming, two hypotheses have been dominant, suggesting respectively that the left-hemisphere/fight-hand system is less variable in its motor output (e. g., Annett, Annett, Hudson, & Turner, 1979; Roy & Elliott, 1986, 1989), or mediates more efficient execution of error corrections utilizing sensory feedback (e.g., Flowers, 1975; Todor & Doane, 1978, Todor & Cisneros, 1985). There is now considerable evidence, derived from experiments in which vision has been manipulated directly

Correspondence to: R. G. Carson now at Department of Human Movement Studies, University of Queensland, Brisbane, Qld 4072, Australia

1 We are grateful to an anonymous reviewer for clarification of this distinction.

(e.g., Carson, Chua, Elliott, & Goodman, 1990; Roy & Elliott, 1986, 1989), to refute an explanation based on the use of visual feedback. It is possible to identify two positions within the class of variability explanations3 Roy and Elliott (1986, 1989) have proposed that the basis of the preferred-hand advantage may be the ability of the right-hand system to modulate force more efficiently (see also Peters, 1980). In this scheme, it is supposed that the left hand becomes proportionately more variable in generating force as the absolute level of force is increased. As such, differences between the hands may be represented as differences in the slopes of force - force-variability curves. One may also conceive of differences in the intercepts of these curves. Annett et al.' s (1979) account does not explicitly treat the distinction between slope and intercept. Rather, they suggest that the output of the left hand is "simply more variable" than that of the fight hand, the implication being that the difference in variability between the hands is independent of the required level of force. Given that for any specific movement goal the impulses produced over a course of trials vary, Schmidt, Zelaznik, Hawkins, Frank, and Quinn (1979) hypothesized that the variability in impulse size can be decomposed into two separate and independent sources. These are variability in the duration over which the impulse is described, and variability in the force component of the impulse. This model, together with experimental data, suggest a roughly linear relationship between the magnitudes of the force and time components of the impulse and their respective variabilities (see Schmidt, Sherwood, Zelaznik, & Leikind, 1985, for a critical review). Thus, in idealized cases, when the duration of the impulse is held constant, doubling the magnitude of the force component of the impulse would double the amount of force variability. Similarly, as the duration of the impulse is increased by a factor of 2, the portion of impulse variability attributable to variations in time is increased twofold. As variability in the muscular impulses are thought to be a major determinant of accuracy, it is therefore predicted that larger impulses wiI1 lead to reduced terminal accuracy (Schmidt et al., 1979).

292

Although both Roy and Elliott (1986, 1989) and Annett et al. (1979) subscribe to the impulse-variability hypothesis (Schmidt et al., 1979), Roy and Elliott (1989) have been more explicit as to the consequences of increased impulse requirements predicted with regard to the magnitude of performance asymmetries. They suggest that the hands are distinguished by differences in the extent to which impulse variability scales with impulse magnitude, and that asymmetries would be accentuated for movements that require impulses of greater magnitude. While Roy and Elliott (1989) failed to confirm this hypothesis on the basis of outcome measures, their manipulation of the impulse requirements (by movement time and movement amplitude) was somewhat indirect. Nor did they examine the underlying characteristics of the movements reflected in their kinematics. In this study, the muscular impulse required to produce an aiming movement was manipulated by the addition of mass to the limb in circumstances in which subjects were required to move both as quickly and as accurately as possible within a specified criterion movement time. In Experiment 1, the required resultant impulse size was also manipulated by variation of movement amplitude (cf. Roy & Elliott, 1989). Kinematic recording techniques were employed to derive resultant impulse characteristics. The time required to reach peak velocity provided a measure of the duration of the initial impulse, whereas variability in the time to reach peak velocity indexed variability in the time course over which the impulse was described. Peak velocity reflected the magnitude of the resultant impulse, while variability in peak velocity revealed the associated variability. The derivation of standard performance measures allowed us to examine the relationship between impulse characteristics and movement outcomes, and the extent to which these relationships differed between the hands.

Two WATSMART cameras were positioned 2.4 m apart - 2.3 m from the ground, and at a perpendicular distance of 1.4 m from the digitizing tablet - with the respective viewing fields converging between the subject and the tablet. During practice trial blocks, it was ascertained that the stylus tip could be seen by both cameras when located at the starting position and at the termination of the movement. The camera positions were calibrated immediately before each experimental session (calibration errors . 10). Kinematic measures. The kinematic variables analyzed in

this experiment were peak velocity, variability in peak velocity, time to peak velocity, variability in time to peak velocity, and time from peak velocity until the end of the movement. All the kinematic variables were analyzed sep-

Table 2. Experiment 1: Radial error (ram), constant error (mm), and variable error (ram) as a function of weight, hand, and movement distance No Weight R.E.

Light C.E.

Heavy

V.E.

R.E.

C.E.

V.E.

R.E.

C.E.

V.E.

Hand Left Right

9.26 8.82

0.81 0.00

7.71 7.35

9.91 8.71

0.70 0.09

8.38 7.48

9.38 9.66

0.69 -0.45

8.35 8.75

Distance 30 cm 40 cm

8.41 9.67

1.17 -0.36

7.52 7.53

8.63 9.99

1.22 -0.43

7.59 8.27

8.93 10.1

0.78 -0.54

8.01 9.09

294 Table 3. Experiment 1: Numerical results ~om ANOVAs (kinematic measures) Effect

F-test

Peak Velocity (PV) Weight Condition Distance

F(2, 14) = 12.55 F(1, 7) = 34.95

p