This article was downloaded by:[BIUS Jussieu/Paris 6] [BIUS Jussieu/Paris 6] On: 6 June 2007 Access Details: [subscription number 772812101] Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Ergonomics

Publication details, including instructions for authors and subscription information: http://www.informaworld.com/smpp/title~content=t713701117

Movement control characteristics of aiming responses a

Les G. Carlton a Motor Behavior Laboratory, Institute for Child Behavior and Development, University of Illinois at Urbana-Champaign. Champaign, Illinois. U.S.A

To cite this Article: Carlton, Les G. , 'Movement control characteristics of aiming responses ', Ergonomics, 23:11, 1019 - 1032 To link to this article: DOI: 10.1080/00140138008924811 URL: http://dx.doi.org/10.1080/00140138008924811

PLEASE SCROLL DOWN FOR ARTICLE Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf This article maybe used for research, teaching and private study purposes. Any substantial or systematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material. © Taylor and Francis 2007

Downloaded By: [BIUS Jussieu/Paris 6] At: 10:09 6 June 2007

ERGONOMICS,

1980, VOL. 23, NO. 11, 1019-1032

Movement control characteristics of aiming responses* By LES G. CARLTON Motor Behavior Laboratory, Institute for Child Behavior and Development, University of Illinois at Urbana-Champaign, Champaign, Illinois 61820, U.S.A.

Two experiments were conducted investigating the movement patterns produced in the completion of aiming responses. Movement displacement, velocity and acceleration patterns were examined in the first experiment in an attempt to determine the control processes used in discrete, peg transfer and reciprocal tapping tasks. The kinematic parameters indicated that each of these tasks were characterized by discrete error corrections occurring near the target. Experiment 2 demonstrated that under high index of difficulty conditions responses are characterized by multiple discrete corrections designed to eliminate the discrepancy between the position of the hand and the target. These findings are discussed in relation to a discrete feedback interpretation of Fitts' law.

I. Introduction Even with the recent growth of interest in motor control processes, the work of Fitts (Fitts 1954, Fitts and Peterson 1964) still remains as one of the few precise laws describing motor performance (Kerr 1978). Fitts (1954) quantified the relationship among movement speed, amplitude and accuracy in aiming responses, showing that movement time (MT) could be predicted from the formula MT=a+blog 2 (2Ajw), where A is the amplitude of the movement, W is the width of the target and a and bare constants. This formula predicts linear increases in MT with increases in the index of difficulty of the task; the index of difficulty (ID) being defined as, ID = log, (2AjW). Adjustments to' Fitts' law have been suggested (Welford 1968) but in general the law tends to hold in a variety of experimental situations (for example, Fitts and Peterson 1964, Kerr 1973, Langolf et al. 1976) and across a broad perspective of populations (for example, Welford et al, 1969, Wade et al. 1978). Various control processes have been hypothesized in order to account for this lawful relationship with the most accepted explanations centred around feedback models. Crossman and Goodeve (1963) utilizing linear feedback theory showed that a sampling-proportional control feedback model leads directly to Fitts' law with feedback control being intermittent in nature. Keele (1968) later proposed an alternative feedback control system with movement control relying heavily on visual feedback information. This model assumes that movements toward a target consist of an initial movement and as many corrective movements, based on visual feedback information, as are necessary for the target to be contacted. These feedback explanations have recently been criticized (Schmidt 1976, Schmidt et al. 1978) on the grounds that the time to obtain and process error information may be prohibitively long for feedback information to be utilized for the control of rapid responses. Schmidt et al. (1978) suggest instead that the speed-accuracy relationship can be explained by a • Requests for reprints should be addressed 10 Les G. Carlton, Department of Health, Physical Education and Recreation, University of Houston, Houston, Texas 77004, U.S.A. O()I~-OIW/IIO/2JJ J

1019 SHill

l)

1980 Taylor & Francis Ltd

Downloaded By: [BIUS Jussieu/Paris 6] At: 10:09 6 June 2007

1020

L. G. Carlton

movement output variability model. This model postulates that the output of the motor system contains 'noise' represented as within subject variability, and this variability is related to the amplitude and the speed of the movement in a way consistent with Fitts' law. Some insight as to the control process used in the production of aiming responses, and the merit of the alternative explanations of the speed-accuracy relationship described by Fitts, can be gained .by examining the movement patterns produced in the completion of aiming responses. According to feedback theories aiming movements are made up ofa number of discrete sub-movements each of which have the same relative accuracy. An aiming movement requiring some precision might then be made up ofa number of movements represented by acceleration and deceleration phases. An alternative form of movement control (Annett et al, 1958) suggests that aiming movements are made up of two phases. The first phase consists of the initiation of the response while the second phase represents the subject actively monitoring his movement. The secondary monitoring phase has also been suggested by Pouiton (1974) who hypothesized that in the secondary phase the target is approached at a constant slow rate with the decision to stop made one reaction time before the target is reached. Movement accuracy would then be proportional to the distance remaining to the target when the decision to stop is made. A recent study of movement patterns (Langolf er al. 1976)shows mixed support for a discrete feedback correction hypothesis. Langolf et al.; using somewhat more elaborate recording techniques than earlier experiments, found that very short movements in a peg transfer task performed under a micrscope were discontinuous. The initial or primary movement lasted approximately 200ms and was followed by a discrete corrective response. Longer movements using a reciprocal tapping task, however, did not show breaks in the movement pattern. The effect of increasing the ID of the task was to slow down the movement in general. Even though some discrepancies exist, Langolf et al . (l976) concluded that the "visually mediated discrete-correction model appear to be [a] feasible qualitative explanation of movement control" (p. 127). Although many answers concerning the control of aiming movements may lie in the study of movement patterns, previous investigations have yielded equivocal results. Studies show that discrete corr-ections do (for example, Crossman and Goodeve 1963) or do not (for example, Langolf et al, 1976) occur in aiming responses, and if they do occur they may be represented as further discrete movements (Carlton 1979) or as a homing phase (Annett er al. 1958). Some of these discrepancies may be due to the varying types of tasks investigated or to a lack of sophistication in measuring techniques. From the data presently available, specific explanations for the speedaccuracy relationship described by Fitts' law are not available.

2. Experiment 1 The present experiment was designed to determine whether discrete corrective movements are characteristic of reciprocal tapping, peg transfer, and discrete aiming responses. By measuring displacement patterns 'using rapid sampling techniques some indication of the underlying control processes predominantly used in the production of aiming responses may be obtained. 2.1. Method 2.1.1. Subjects. Subjects were three right-handed volunteers from the University of lIIinois who.had no previous experience with the experimental tasks.

Downloaded By: [BIUS Jussieu/Paris 6] At: 10:09 6 June 2007

Mooement control characteristics of aiminq responses

1021

2.1.2. Apparatus and procedures. Subjects participated in each of three accuracy tasks (discrete aiming, reciprocal tapping and peg transfer) with control processes and error corrections examined at two levels oflD. Subjects completed trials on two consecutive days. The first day consisted of practice trials at each of the six conditions. A second practice phase and the test trials were completed on day 2. The practice phase consisted of 50 trials for the two discrete conditions and four trials for the reciprocal tapping and peg transfer tasks. The order of presentation for the six conditions in both the practice and test phase was determined randomly for each subject. The specific procedures for each of the three tasks are presented separately. 2.1.3. Discrete task. The discrete MT apparatus consisted of a 75 x 18 ern aluminium plate with a defined start and a hole in which a target could be placed. The centre of the target, which consisted of a copper disc 1·59 or O' 31 cm in diameter, was placed 20·3 cm from the start position. This resulted in IDs of 4·65 and 7·00 respectively. A red 28 V incandescent lamp, 0·9 cm in diameter, was located beside the target. Movement time and errors were automatically recorded on a Hewlett-Packard printer via a digital interface unit. Movement time was the time interval between the stylus releasing from the start position to contact with the target. Movements which did not come into contact with the target but which resulted in the stylus coming into contact with the surrounding area were recorded as errors. The order of presentation of the two ID conditions was determined randomly for each subject with all trials at the initial ID condition being presented before testing began at the second ID. Subjects received 80 trials at each of the two ID conditions. The first 50 trials were practice trials and the last 30 trials were test trials. Movement patterns were determined from eight of the test trials with the use of a high speed cinematography system. The eight film trials were selected randomly from the 30 test trials. All movements were made in a right to left direction. 2.1.4. Fitts' tapping task. The tapping apparatus was similar to that used by Fitts (19541 in his original investigation of aimed movements. Two target plates, 15·27 ern long and either 1·59 or 0·31 em wide, were mounted on a soft base. The soft base extended the length of the targets and 3 em on either side of the target area. These 'error' plates (the soft bases) were wide enough to record all movement errors. The 20·3 ern distance paired with the 1·59 and O' 31 em target produced IDs of 4·65 and 7·00 respectively. An impulse counter automatically recorded the number of hits, and errors were determined from marks produced on the soft base. Each series of movements started in a neutral position with the stylus halfway between each target. Atthe presentation ofa visual signal the subject was to begin alternately tapping the two target plates as rapidly as possible without touching either of the error plates. Subjects were encouraged to move as rapidly as possible but were told to emphasize accuracy. The presentation order of the two ID conditions was determined randomly for each subject. Tapping trials lasted 15s with a 60 s rest between trials. Subjects received four practice trials followed immediately by three test trials. The practice and test trials were identical with the exception that movement patterns were recorded from the test trials. 2.. 1.5. Peg transfer, The peg transfer apparatus consisted of a peg O' 30 ern in diameter and 4 ern long and two boards. Two aluminium blocks 5-4 ern long x 3-8em wide x 1·27em deep were mounted on each board. A hole I cm deep was drilled in the centre of each block with a peg clearance of either 1·59 or 0·31 em, The blocks were placed so

Downloaded By: [BIUS Jussieu/Paris 6] At: 10:09 6 June 2007

1022

L. G. Carlton

that their centre to centre distance was 20·3 cm. Pairing the two 1·59 cm blocks and the two O· 31 cm blocks at this movement distance resulted in the same IDs as those used in the two previous tasks, 4·65 and 7·00. The procedures were similar to those used in the Fitts' tapping task. Subjects were required to move the peg back and forth between the holes as rapidly as possible. The order in which the subjects received the [D conditions was determined randomly and the presentation of the trials was the same as that used in the tapping task. In both the peg transfer and reciprocal tapping tasks portions of the lirsl and third test trial were filmed. The camera was started after approximately 8 s of the trial had been completed. This yielded eight filmed responses for each combination of movement task and [D. A high speed cinematography system was used to record the movement patterns of the lilmed responses. The subject was seated at a table and the movement apparatus was placed so that the plane of motion was directly in front of the subject. A 16 mm battery driven d .c. model (LoCam) camera secured to a tripod was used for all filming. Attached to the camera was a timing light generator which was used to drive lightemitting diodes within the camera. The timing light generator was set at 100 Hz, and provided !l very accurate time base on the film from which the speed of the camera was ascertained. A 50 mm lens was used and the camera was placed 1·37 cm from and perpendicular to the plane of motion. Responses were filmed at a rate of222 frames per second.

2.1.6. Data allalysis. For all three tasks, the overall MTs for each movement were obtained from the movement pattern output. From the film representation movement displacements were measured with the aid ofa Vangard Motion Analyser. The location of the stylus was determined in X (horizontal) and Y(vertical) coordinates from which the movement displacement, velocity and acceleration patterns were determined for the X and Yaxes separately. The resultant movement patterns were also calculated from these measures. Smoothed displacement, velocity and acceleration data were obtained from the cubic spline method (McLaughlin et al. 1977) using an estimated measurement error of 0·2mm. These kinematic patterns were used to examine the control processes employed in the production of aiming movements and to locate the presence of any discrete movement corrections. Crossman and Goodeve (1963) ,have shown that responses where corrective movements are not attempted are characterized by uniform increases and decreases in velocity and symmetrical acceleration and deceleration phases. Thedeceleration phase in these non-corrected responses showed a smooth approach to zero acceleration with no abrupt changes in the deceleration values or accelerations toward the target. The primary criterion adopted for determining the occurrence ofa corrective movement in the present experiment was the acceleration of the stylus as it approached the target area. This secondary acceleration, was associated with the initiation of a movement command intended to correct a discrepancy between the position of the stylus and the target. The initiation of a corrective response was also associated with abrupt uncharacteristic decelerations which took place near the target. In the case of accelerations near the target, corrections were defined by the shift from deceleration to acceleration or by the turning point in the velocity curve where decreases in velocity changed to increases. Uncharacteristic decelerations near the target were defined as those where a turning point took place in the deceleration portion of the curve. That is, where deceleration values were approaching zero and were followed by a change to increases in the rate ofdeceleration.

Downloaded By: [BIUS Jussieu/Paris 6] At: 10:09 6 June 2007

Ion

Movement control characteristics oj aiminq responses

2.2. Results and discussion The mean MTs for the three tasks and two ID conditions are presented in table I. The peg transfer and reciprocal tapping tasks were further separated according to their movement direction with movements being produced in a left to right or right to left direction. The MTs for the peg transfer and reciprocal tapping task represent the actual time to complete the response and excludes the time the stylus was in contact with the target. Table 1 shows the large increases in MT for the high ID condition as well as differences in MTs for the three tasks. In general the reciprocal tapping task was completed more slowly than the other two tasks with the discrete task being completed most rapidly. In addition, movements in the right to'left direction were completed more quickly than movements from left to right. This may have been due to the hand partially obstructing vision of the target in left to right movements. The findings of primary interest were those which pertained to the control processes used in the production of the various types of aiming responses. Examination of the kinematic parameters revealed that responses could be characterized as containing no discrete corrections, as clearly having one or more discrete corrections or as having some characteristics of each of these control processes. Figure I represents the resultant velocity and acceleration pattern from an individual trial. This trial was performed on the discrete task at the 4·65 ID condition and is representative of responses where no apparent corrections took place. The resultant velocity pattern, which depicts the actual velocity of the stylus taking into account both the vertical and horizontal dimensions, is represented as an increase in velocity from the start of the movement with peak velocity occurring at approximately 115 ms. The stylus then slowed until the target was contacted at 320 ms. The acceleration pattern shows the nearly symmetrical acceleration and deceleration phases with no indication of discrete corrections as the stylus approached the target. The movement pattern characteristics of responses which lacked corrections were in contrast to those where discrete corrections appeared to take place. Figures 2 and 3 depict the movement patterns produced in an individual reciprocal tapping trial made at the 7·00 ID condition. The vertical displacement and velocity patterns are presented in figure 2. The displacement pattern displays the initial movement upward with the stylus reaching its peak height at 250ms and then proceeding down toward the target. At about 500 ms a levelling-off period occurred indicating that the movement Table I. Mean MTs for discrete, peg transfer and reciprocal tapping tasks as a function of ID and movement direction (ms) based on three subjects and 144 total trials (48 trials per task).

ID Directional

Discret!' Right to left Peg transfer Left to right Right to left Reciprocal tapping Left to right

4·65

7·00

mean

351 342

534 675

509

363 381

708 692

536 537

431

816

624

Mean 443 522

580

Downloaded By: [BIUS Jussieu/Paris 6] At: 10:09 6 June 2007

· Vi :

:!O -J

~ ~

W

>

• J

• • J

J

B 0 0. 00

,.~

..

,

UO- I

'.IS

~

~ B

Vi

• J

"">:

V>

:!O u u

a:

s.u

.. "

no·\

•• TIME lSI

Figure I.

Resultant velocity and acceleration patterns for a discrete response showing no corrections.

,m

TIME(S(

Figure 2.

Vertical displacement and velocity for a reciprocal tapping movement showing a

discrete'correction.

Downloaded By: [BIUS Jussieu/Paris 6] At: 10:09 6 June 2007

Movement control characteristics of aiminq responses

1025

J!?

"

~ (f)

~

: