Reaction time latencies of eye and hand movements in single- and dual-task conditions.

Abstract

The goal of this study was to investigate whether ocular and hand motor systems operate independently or whether they share processes. Using dual-task methodology, reaction time (RT) latencies of saccadic eye and hand motor responses were measured. In experiment 1, the hand and eye motor systems produced rapid, aimed pointing movements to a visual target, which could occur either to the left or right of a central fixation point. Results showed that RT latencies of the eye response were slower in the dual-task condition than in the single-task condition, whereas the RT latencies of the hand response were virtually the same in both conditions. This interference effect indicated that the ocular and manual motor systems are not operating independently when initiating saccadic eye and goal-directed hand movements. Experiment 2 employed the same experimental paradigm as experiment 1, except for one important modification. Instead of a goal-directed hand movement to the target stimulus, subjects had to make a button-press response with either the index or middle finger of the right hand dependent upon whether the stimulus occurred to the right or left of the control fixation point. The aim of experiment 2 was to investigate the issue whether the observed interference effect in experiment 1 was specific or non-specific (e.g. overhead costs due to coordinating any two responses). The finding that saccadic eye movements and button-press responses in the dual-task condition could be initiated without delay relative to the single-task conditions, supports the specific interference interpretation.

Full text

Page 1

Exp Brain Res (1994) 97:471476
Experimental
Brain Research
y Springer-Verlag 1994
Reaction time latencies of eye and hand movements
in single- and dual-task conditions
Harold Bekkering 1, Jos J. Adam 1, Herman Kingma 2, A. Huson 1, H.T.A. Whiting 3
1 Department of Movement Sciences, University of Limburg, 6200 MD Maastricht, The Netherlands
2 Division of Balance Disorders, Academic Hospital of Maastricht, 6200 MD Maastricht, The Netherlands
3 Department of Psychology, University of York, UK
Received: 10 May 1993 / Accepted: 7 September 1993
Abstract. The goal of this study was to investigate
whether ocular and hand motor systems operate inde-
pendently or whether they share processes. Using dual-
task methodology, reaction time (RT) latencies of sac-
cadic eye and hand motor responses were measured. In
experiment 1, the hand and eye motor systems produced
rapid, aimed pointing movements to a visual target,
which could occur either to the left or right of a central
fixation point. Results showed that RT latencies of the
eye response were slower in the dual-task condition than
in the single-task condition, whereas the RT latencies of
the hand response were virtually the same in both condi-
tions. This interference effect indicated that the ocular
and manual motor systems are not operating indepen-
dently when initiating saccadic eye and goal-directed
hand movements. Experiment 2 employed the same ex-
perimental paradigm as experiment 1, except for one im-
portant modification. Instead of a goal-directed hand
movement to the target stimulus, subjects had to make a
button-press response with either the index or middle
finger of the right hand dependent upon whether the
stimulus occurred to the right or left of the control fixa-
tion point. The aim of experiment 2 was to investigate the
issue whether the observed interference effect in experi-
ment 1 was specific or non-specific (e.g. overhead costs
due to coordinating any two responses). The finding that
saccadic eye movements and button-press responses in
the dual-task condition could be initiated without delay
relative to the single-task conditions, supports the
specific interference interpretation.
Key words: Eye-hand coordination - Reaction times -
Motor control - Dual-task methodology - Human
Correspondence to: H. Bekkering, e-mail: Harold Bekkering @ BW.
Rulimburg. NL
Introduction
This paper focuses on temporal coupling between eye
and hand movements in the context of pointing to a visu-
al target. In the last decade two relevant questions re-
garding this issue have been examined: first, whether ini-
tiations of eye and hand responses occur in a fixed or
variable order (e.g. Abrams et al. 1990; Biguer et al. 1982;
Prablanc et al. 1986) and, second, whether eye and hand
responses are initiated by one common command signal
or by different command signals (e.g. Fischer and Rogal
1986; Frens and Erkelens 1991; Herman et al. 1981; Pail-
lard (1982); Prablanc et al. 1979).
Results of studies addressed to the first question are
straightforward. Almost always the eyes start moving to-
ward the target before the hand, and because eye move-
ment durations are quite brief, the eyes usually arrive at
the target before the hand starts to move (Abrams et al.
1990).
Answers to the second question, however, are not
quite so straightforward. Typically, a correlational ap-
proach has been used in order to determine the degree of
co-variation between reaction time (RT) of eye and hand
movements, high correlations being interpreted as evi-
dence that eye and hand responses are initiated by a
common command signal, low correlations as evidence
for independent command signals.
Experiments using the correlational method have
yielded inconsistent results, widely differing correlations
having been reported both within and between studies.
For example, low correlations (r < 0.4) were reported by
Biguer et al. 1982, and by Frens and Erkelens 1991 ; mod-
est correlations (0.4<r<0.6) by Gielen et al. 1984,
Prablanc et al. 1979, and by Fischer and Rogal 1986; high
correlations (r>0.6) were found by Herman et al. 1981
and also in some of the conditions of experiments by
Frens and Erkelens 1991 and by Fischer and Rogal 1986.
Because most of the studies fail to report high correla-
tions, and because low and modest correlations can easi-
ly be attributed to a shared perceptual processing corn-

Page 2

472
ponent (Gielen et al. 1984), strong evidence in support of
the common command hypothesis is lacking.
Failure to obtain evidence for the common command
hypothesis, however, does not automatically constitute
evidence in favour of the independent commands hy-
pothesis. Theoretically it is still possible that eye and
hand motor systems do work together, but that the de-
gree of cooperation is insufficient to realize strong corre-
lations. To investigate the issue whether eye and hand
motor systems operate independently or whether they
share processes, dual-task methodology 1 was employed
in this study. In the single-task condition subjects had to
make either eye or hand movements to a visual target,
while in the dual-task condition both hand and eye were
required to move to the visual target.
The logic behind dual-task methodology is as follows.
When eye and hand movements are executed simulta-
neously, without an increase in RT latencies in the dual-
task condition relative to the single-task conditions,
strong support is found for the assumption that both
responses are mediated by independent processes. In the
case that the RT latencies in the dual-task condition dif-
fer from those in the single-task conditions, support is
found for the notion that movements produced by eye
and hand motor systems share processes.
Experiment 1
Experiment 1 was designed to study RT latencies of eye
and hand movements in single- and dual-task conditions.
The experimental task required subjects to move (eyes
alone, hand alone, or eyes and hand concurrently) as
quickly as possible to a large target stimulus appearing
either to the left or right of a central fixation point. The
choice of a large target size was motivated by the wish to
impose low accuracy constraints, i.e. to demand only a
ballistic hand response without the need for visually
guided error corrections.
Materials and methods
Subjects
Twelve right-hand preferent undergraduate students (three male
and nine female, mean age 22 years), participated as non-paid, vol-
untary subjects. Nine subjects had normal vision, while the three
subjects who normally wore glasses were not allowed to wear them
during the experiment. As all these subjects had an acuity of at least
18/20 D, the nature of any impairment they experienced was incon-
sequential for performing visual tasks at short distances as in the
present task. All were naive as to the purpose of the experiment.
Design
Each subject served in three conditions on a single day, with each
condition lasting 10 min with a break of 5 rain between conditions.
The term dual-task methodology in this paper refers to the proce-
dure of studying RT latencies of eye and hand movements in single
and dual task conditions. Note that the dual-task condition is used
in most studies on eye-hand coordination but the single-task condi-
tion is not
In the eye-only condition, only the eyes moved to the stimulus; in
the hand-only condition, only the hand moved to the stimulus while
the eyes remained fixated on the fixation point; in the eye-and-hand
condition, both eyes and hand moved to the stimulus. Order of
conditions was balanced between subjects.
Apparatus
The experiment was carried out in a normally illuminated room
which contained many visual cues. The subject stood behind a
horizontally placed transparent X-Y tablet (Scriptel RDT, 105-
80 cm). Underneath the X-Y tablet a tv-monitor (Phillips, 54-
40 cm) was mounted, controlled by an AT-MSDOS computer. On
the monitor a fixation sign was presented in line with the subject's
eyes at a distance of about 28 cm. The fixation point was a 'filled
dot', diameter 2 deg of arc, and the stimulus was a 'filled rectangle'
presented at a distance equivalent to 15.9-24.9 deg of visual angle.
Positions of subjects' eyes were monitored with an IRIS system.
This system records eye position by reflection of iris-sclera
boundaries by means of infrared light; the output of the IRIS sys-
tem is low-pass filtered (200 Hz, --3 dB); for details see Reulen et al.
(1988). The analog output from the IRIS was digitized at a rate of
200 Hz. Calibration was performed at the beginning of each session.
The head was fixated by a frame which was connected to the IRIS
35 cm above the middle of the X-Y tablet, so that the eyes were
directly above and looking down at the control fixation dot.
The subjects were requested to move a stylus with their right
hand over the surface of the X-Y tablet which recorded X-Y coor-
dinates with an accuracy of 0.1 mm at a rate of 172 Hz.
Procedure
A series of 10 practice trials preceded the 20 experimental trials. At
the beginning of each trial the subjects were requested to place the
stylus on the fixation point. After 2 s, the target stimulus appeared
randomly either to the left or to the right of the fixation point.
Subjects were instructed to make saccadic eye movements and/or
ballistic hand movements towards the target. They were also in-
structed to minimize both RT and movement time (MT) of eye
and/or hand.
Eye-movement analysis
The position data generated by the IRIS system were analysed off
line by means of the software program NYSTAGLINER (Toennies,
Wfirzburg) to identify the occurrence of saccades. The beginning of
a saccade was defined when the signal exceeded the value of the root
mean square (RMS) of the previous 120 ms with a factor 3 within a
subsequent window of 20 ms. A trial was excluded from the analysis
when another saccade or eye blinking appeared within 100 ms be-
fore or after this point of detection. The same criteria were used to
check fixation of the eyes in the hand-only condition. Because all
subjects showed almost perfect binocular coordination between
horizontal saccades (correlation values for each subject >0.99), on-
ly data concerning the right eye are presented 2
Hand-movement analysis
Detection of changes in hand movements (RTs) were calculated by
a computer algorithm which included four criteria. The first three
2 All trials were visually inspected; in about 10% of the trials of
both eye-only and eye-and-hand conditions the automatic detection
of a saccade was manually corrected. These manual corrections
were invoked when significant eye drifts occurred

Page 3

samples had to differ at least one unit in x-direction (one unit
corresponds to 0.1 mm) from each other, the fourth more than one,
the fifth more than two and the sixth sample had to differ more than
three units from its preceding x-coordinate.
Correlation analysis.
Within-subject Pearson correlation coefficients were calculated on a
trial by trial basis between eye and hand RTs in the dual-task
condition. In addition, a between-subject Pearson correlation was
calculated using the subjects' mean RTs of eye and hand in the
dual-task condition.
Results
The results of one subject were excluded because eye
blinks occurred in more than 25% of the trials which
required eye movements. For the remaining 11 subjects,
10.0% of trials which required saccadic eye movements
Table 1. Mean RT latencies (ms) and standard deviatons (ms) for eye
and hand aiming responses in single- and dual-task conditions (ex-
periment 1)
Eye Hand
Single-task
Dual-task
222 + 35
261 ___41
234 + 41
238 _+42
AIMING RESPONSE
RT (ms)
35O
300
250
200
Y
EYE
HAND
150 I I
SINGLE DUAL
TASK CONDITION
Fig. 1. Mean RT latencies (ms) of eye and hand aiming responses in
both single- and dual-task conditions (experiment 1). Saccadic eye
RT latencies are longer in the dual-task condition than in the sin-
gle-task condition
473
in the eye-only condition and 11.8% in the eye-and-hand
condition were excluded (see Materials and methods Eye
movement analysis).
Table 1 shows mean RT and standard deviation of eye
and hand movements in single and dual-task conditions.
An ANOVA yielded a significant main effect for task
condition (single vs dual), F(u0)= 12.7, p<0.01, indicat-
ing longer reaction times in the dual-task than in the
single-task condition (249 and 229 ms, respectively). The
main effect of type of movement (eye vs hand) was not
significant, F(uo~ = 0.1, p > 0.2. Importantly, a significant
interaction effect was found between type of movement
and task condition, F~l,m)= 19.3, p<0.01. A paired, two-
tailed t-test indicated that RT of the eyes was significant-
ly longer (39 ms) in the dual-task condition than in the
single-task condition (T(m ~ = 4.25, p < 0.002).
RT latencies of the hand were virtually identical in
single and dual-task conditions (Too) = ~0.23, p > 0.8). The
RT latencies of the eyes and the hand were not signifi-
cantly different from each other in the dual-task condi-
tion (Too ~ = 1.45, p > 0.15). 3 None of the individual (with-
in-subjects) correlations between eye and hand RTs was
significant, all p values being larger than 0.05. The be-
tween-subjects correlation was r = 0.14, p > 0.05.
Discussion
The main finding of experiment 1 was that RT latencies
of the eyes were consistently and significantly longer in
the dual-task condition than in the single-task condition,
while those of the hand did not differ significantly. In
addition, no significant correlations were found. Non-sig-
nificant or low correlations do not support the common
command hypothesis but, as stated in the Introduction,
low correlations do not constitute strong evidence for the
notion that the eye and hand motor systems operate in-
dependently. In fact, the interference effect for the eye
response in the dual-task condition indicates that the oc-
ular and manual motor systems do not operate indepen-
dently; rather, at some stage, processes seem to be
shared.
The question to which this finding gives rise is whether
or not the observed interference effect is specific or non-
specific i.e. the consequence of being required to produce
any two responses, or the sharing of specific processes in
the context of pointing to a target.
One way to discriminate between these two interpre-
tations is to manipulate the nature of the manual re-
sponse. This is done in experiment 2.
Experiment 2
The same paradigm as for experiment 1 was used except
for one important modification. Instead of a goal-direct-
ed hand movement to the target stimulus, subjects had to
make a button-press response with either the index or
3 The mean difference between RT latencies of eye and hand in
dual-task condition was 23 ms. However, the large variation in eye
and hand RTs (SD 54) makes this difference non-significant

Page 4

474
middle finger of the right hand. Specifically, the left target
stimulus required a button-press response with the index
finger and the right target stimulus a button-press re-
sponse with the middle finger. The same saccadic eye
movement was required as in experiment 1.
The purpose of this manipulation is straightforward.
If experiment 2 shows the same interference effect as
found in experiment 1, support is provided for non-
specific interference. On the other hand, if the interference
effect disappears, support is found for a specific interfer-
ence effect caused by a sharing of processes associated
with the control of coordinated aimed eye and hand
movements.
Materials and methods
Subjects
Twelve right-hand preferent undergraduate (five male and seven
female) students (mean age 23 years) participated as non-paid, vol-
untary subjects. All had normal vision, and all were naive as to the
purpose of the experiment. None of them had taken part in the
previous experiment.
Procedure
Instead of making a goal-directed hand movement, subjects were
requested to press one of two buttons on a response box with either
the index finger, when the stimulus was presented to the left of the
fixation sign, or with the middle finger when the stimulus was to the
right. The subjects were required to keep the two fingers on the
buttons during the experiment.
The design, apparatus and eye-movement analysis were the
same as in experiment 1.
BUTTON-PRESS RESPONSE
RT (ms)
350
300
250
200
FINGER
EYE
150 i i
SINGLE DUAL
TASK CONDITION
Fig. 2. Mean RT latencies (ms) of eye and button-press responses in
both single- and dual-task conditions (experiment 2). Saccadic eye
RT latencies and button-press RT latencies are similar in single and
dual-task conditions
The most important result of experiment 2 was the
absence of a significant interaction effect between type of
movement (eye vs finger) and task condition (single vs
dual), (F(1,H) = 0.6, p > 0.2).
Finger-movement analysis
A connection was made between the response box and the AT-MS-
DOS computer by means of the joystick port. This port scanned
both button states (high or low value) at a rate of 1000 Hz. The
values were analysed off-line. A trial was excluded when no re-
sponse occurred within a period of 750 ms or when a subject
pressed the wrong button.
Results
We excluded 3.1% of trials requiring saccadic eye move-
ments in both the eye-only and the eye-and-finger condi-
tion, 0.7% of trials requiring finger responses in the fin-
ger-only and 1.7% of trials in the eye-and-finger condi-
tion.
Figure 2 shows the mean RT latencies of eye and fin-
ger responses in single- and dual-task conditions. A main
effect was found for type of movement, indicating that the
RT latencies of the eyes (217 ms) were significantly short-
er than the RT latencies of the finger (316ms),
(F(1,11)=44.5, p<0.001). No main effect was found for
task condition, (F(1,~l~ = 3.2, p> 0.05).
Discussion
The purpose of experiment 2 was to investigate whether
the interference effect between the eye and hand response
observed in experiment 1 was due to specific or non-
specific interference. The finding that saccadic eye move-
ments and button-press responses in the dual-task condi-
tion could be initiated without delay relative to the sin-
gle-task conditions, supports the specific interference in-
terpretation.
General discussion
The experiments reported here produced three major
findings. Experiment 1 showed the absence of a temporal
order effect. There were no significant differences between
the RT latencies of eye and hand in the dual-task condi-
tion. This experiment also demonstrated an interference
effect. There was an increase in saccadic eye RT latencies
in the dual-task relative to the single-task condition. Ex-
periment 2 showed a task-specific effect i.e. there was no
increase in either the saccadic or button-press RT laten-
cies in the dual-task relative to the single-task condition.

Page 5

475
Temporal order
The common finding in the literature of dual-task condi-
tions involving eye-hand coordination movements to a
target is that the eye starts to move in the direction of the
target before the hand (see Introduction for references).
The discrepant finding reported in experiment 1 seems to
indicate that this order of eye and hand responses may be
due to the specific task requirements. Traditionally, ex-
periments on eye-hand coordination have involved a rel-
atively small target - for example, a light-emitting diode
- which, because of the need for precise foveal informa-
tion to guide the hand accurately to the desired end posi-
tion, may invoke serial order operation of eye and hand
motor systems. In experiment 1, such stringent demands
were not imposed on subjects. Speed and not accuracy of
both eye and hand movements was stressed and, to this
end, large targets were used and only two possible target
positions employed. The absence of a temporal order ef-
fect in the present dual-task condition calls into question
the generality of the eye-first, hand-second phenomenon.
In their first experiment, Abrams et al. (1990) found that
subjects spontaneously executed a saccade toward the
target of their rapid aimed limb movements but that only
three of four subjects tested began to move their eyes first.
They phrased this finding as "The saccade is closely time
locked to the initiation of the limb movement, although
its order of occurrence does not seem crucial: limb move-
ments (e.g., wrist rotation) are equivalent whether they
lead or follow an eye movement" (p. 254). The actual
movement order therefore may depend on the be-
havioural task involved, as for instance the amount of
uncertainty about the spatial location of the target.
The finding that saccadic eye movement latencies
were slightly larger in the experiments reported here than
those to be found in the literature can probably be ac-
counted for by the fact that most researchers extin-
guished the fixation light at the moment the target ap-
peared. This extinction has been shown to provide an
extra cue that reduces saccadic latency (e.g. Reulen 1984).
Other investigations within a so-called overlap paradigm
showed saccadic RT latencies of the same order (e.g. Fis-
cher and Rogal 1986).
Task-specific interference effect
The fact that no increase in RT latencies was observed
either for the eye or for the hand in experiment 2, where
finger-press responses were required, suggests that the
observed interference effect of experiment 1 is restricted
to situations in which both hand and eye are functionally
involved in moving to a target position.
In a related study done with eye and hand movements
within a dual-task paradigm, Fischer and Rogal (1986)
found saccadic eye latencies to be independent of
whether or not they were combined with hand move-
ments. However, in contrast to the present study they
used small targets (1 deg x 1 deg). This task constraint
may induce subjects to first move the eyes first to the
target position and subsequently the hand; in other
words, small targets may demand serial operation of ocu-
lar and manual motor systems, while large targets may
demand parallel operation. Frens and Erkelens (1991)
also concluded that, depending on the task constraints,
coordination of hand and eye movements involves paral-
lel and independent control, or the sharing of specific
processes.
Both the finding that eye and hand RTs were not sig-
nificantly different from each other in the dual-task con-
dition and the finding that none of the subjects showed a
significant correlation between eye and hand RTs indi-
cate that the augmentation of saccadic eye RTs in the
dual-task relative to the single-task condition is not due
to an eye-is-waiting-for-the-hand phenomenon. Further
research is being carried out in our laboratory to investi-
gate the nature and locus of the delay that ensues when
saccadic eye responses are combined with goal-directed
hand movements.
Because saccades have stereotyped velocity profiles
and because they show quick orienting behaviour under
all kind of circumstances, the saccadic system has often
been considered to be an automatically reacting open-
loop system (e.g. Carpenter 1988; Posner et al. 1978; Pra-
blanc and Pelisson 1990; Young and Stark 1963); clearly,
the results of experiment 1 in which a significant increase
in saccadic RT latency was observed in the dual-task
condition call the generality of this conceptualization in-
to question.
Acknowledgements. The authors are grateful to Casper Erkelens,
Bernd de Graaf, Jozina de Graaf and two anonymous reviewers
who provided useful comments on an earlier version of this
manuscript. We also wish to thank Iwan the Jong, Huub Jussen,
Hans Konijnenberg, Gerard Simons and Harry Wandler for techni-
cal assistance.
References
Abrams RA, Meyer DEM, Kornblum S (1990) Eye-hand coordina-
tion: oculomotor control in rapid aimed limb movements. J Exp
Psychol 16:248 267
Biguer B, Jeannerod M, Prablanc C (1982) The coordination of eye,
head and arm movements during reaching at a single visual
target. Exp Brain Res 46:301-304
Carpenter RHS (1988) Movements of the eyes. Pion, London
Fischer B, Rogal L (1986) Eye-hand coordination in man: a reac-
tion time study. Biol Cybern 55:253-261
Frens MA, Erkelens CJ (1991) Coordination of hand movements
and saccades: evidence for a common and a separate pathway.
Exp Brain Res 85:682-690
Gielen CCAM, Heuvel van den PJM, Gisbergen van JAM (1984)
Coordination of fast eye and arm movements in a tracking task.
Exp Brain Res 56:154-161
Herman R, Herman R, Maulicci R ((1981) Visually triggered eye-
arm movements in man. Exp Brain Res 42:145-148
Paillard J (1982) The contribution of peripheral and central vision
to visually guided reaching. In: Ingle D, Goodale M, Mansfield
R (eds.) Analysis of visual behaviour MIT Press, Cambridge pp
367-385
Prablanc C, Pelisson D (1990) Gaze saccade orienting and hand
pointing are locked to their goal by quick internal loops. In:
Jeannerod M (ed) Attention and performance XIII. Elsevier,
North-Holland, pp 653-676

Page 6

476
Prablanc C, Echallier JF, Komilis E, Jeannerod M (1979) Optimal
response of eye and hand motor systems in pointing at a visual
target.. Spatio-temporal characteristics of eye and hand move-
ments and their relationship when varying the amount of visual
information. Biol Cybern 35:113-124
Prablanc C, Pelisson D, Goodale MA (1986) Visual control of
reaching movements without vision of the limb 1. Role of retinal
feedback of target position in guiding the hand. Exp Brain Res
62: 293-302
Reulen JPH (1984) Latency of visually evoked saccadic eye move-
ments. I. Saccadic latency and the facilitation model. Biol Cy-
bern 50:251-262
Reulen JPH, Marcus JI, Koops D, de Vries FR, Tiesinga G,
Boshuizen K, Bos JT (1988) The precise recording of eye move-
ment: the Iris technique, part I. Med Biol Eng Comput 26:20-26
Young LR, Stark L (1963) Variable feedback experiments testing a
sampled data model for eye tracking movements. IEEE Trans
Hum Factors Electron 4:38-51