This article was downloaded by: [UZH Hauptbibliothek / Zentralbibliothek Zürich]

On: 01 July 2014, At: 08:36

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Journal of Motor Behavior

Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/vjmb20

Programming Precision in Repetitive Tapping

a a
Denis Glencross & Nicholas Barrett
a
Cognition and Performance Laboratory Department of Psychology, The Flinders University of South Australia
Published online: 13 Aug 2013.

To cite this article: Denis Glencross & Nicholas Barrett (1983) Programming Precision in Repetitive Tapping, Journal of Motor Behavior, 15:2, 191-200, DOI: 10.1080/00222895.1983.10735296

To link to this article: http://dx.doi.org/10.1080/00222895.1983.10735296

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any
purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information.
Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions
 journal of Motor Behavior
1983, Vol. 15, No.2, 191-200
•

Programming Precision in Repetitive Tap.ping •

Denis Glencross
Nicholas Barrett
Cognition and Performance Laboratory
Department of Psychology
The Flinders University of South Australia

ABSTRACT. The present paper reports an experiment using the Fitts' tapping
paradigm. It is concerned with a comparison of movement times and accuracy
during blind and visual repetitive tapping. A blind condition was used to in-
vestigate rapid aiming movements under motor program control, whilst visual
aiming was used to assess the role of visual feedback for control purposes. Sub-
jects in the blind conditions were able to replicate the amplitude specifications of
the task, whereas effective target width was constant for a set amplitude and did
Downloaded by [UZH Hauptbibliothek / Zentralbibliothek Zürich] at 08:36 01 July 2014

not reflect specified target width. Subjects, furthermore, responded more rapidly
when tapping blind. These results are discussed in terms of the magnitude of
forces being attempted as a result of performing a set amplitude, and the role of
visual feedback.
•

A CENTRAL ISSUE in the control of skilled movements is the integra-

tion of sensory information, which is available to the performer during
the movement, with the ongoing control of the movement. In par-
ticular, the integration of visual information is crucial in movements in
which spatial accuracy is important. As Keele (1980) has pointed out, an
actor attempting a movement to reduce the discrepancy between a
visual target and his reaching limb must translate his perception of the
distance to be covered to actual movement. An underlying assumption
is that movement time increases with the distance to be moved and
precision of the outcome, because there are limits to the accuracy of
this translation. The present study examines this translational accuracy
using Fitts' (1954) tapping paradigm for repetitive movements where
visual feedback is either present or absent. In this manner, it was pos-
sible to investigate the relative roles of visual feedback and programmed
control.
Is the accuracy of a speeded movement related to the advanced pro-
gramming of the action or is it dependent upon the quality of feedback

This project was supported by a grant from the Australian Research

Grants Commission. Requests for reprints should be sent to D.}. Glencross, .
Department of Psychology, Flinders University of South Australia, Bed-
ford Park, South Australia, 5042

191
 Denis Glencross and Nicholas Barrett

information available? Crossman and Goodeve (1963) and Keele (1968)

have proposed that the accuracy of an aimed movement is directly
related to the time available to process feedback, and also that the aimed
movement was composed of a series of submovements, each of about
the same duration and relative accuracy, which served to correct ac-
curacy. The duration of each submovement was 200-300 msec, and
these iterative corrections continue until the desired accuracy is attained.
However, a number of alternative strategies have been observed. For
example, Langolf, Chaffin, and Foulke (1976) and Vince (1948) have
shown that, under some circumstances, the whole movement
decelerated as precision increased, allowing for several corrections to
be made. Howarth, Beggs, and Bowden (1971} also report a slowing
down of the entire movement, but with one single correction in the
region of the target. The accuracy of this final correction is related to the
distance from the target at which the final correction occurs.
On the other hand, Megaw (197 4} suggested that for rapid move-
ments the response outcome is directly related to the precision of the
motor program. If this is the case, it is necessary to determine the degr.ee
of precision that can be built into such a program, and how this program
can be amended or modified if the precision is not sufficient to meet the
Downloaded by [UZH Hauptbibliothek / Zentralbibliothek Zürich] at 08:36 01 July 2014

desired outcome or accuracy.

The present study investigated the relative roles of visual feedback
and programmed control in repetitive tapping movements to targets. In
particular, the interest was in comparing the levels of precision achieved
in movements made without visual feedback (viz., blind movements},
to similar movements with visual information. This raises a number of
issues. In rapid blind movements, will programming be related to target
•

size and distance to be travelled? Will the absence of visual feedback

force subjects to adjust response accuracy by varying movement time?
Again, for the same target conditions will precision increase when visual
information is available? What effect will this have on the duration of the
movement?

Method

Subjects. Five female and seven male students or staff, of ages from 17
to 40 years, served as subjects. All were right-handed as assessed by a
verbal questionnaire.
Apparatus and task. The task was a replication of Fitts' (1954} basic
procedure using paper and pencil techniques. Targets were drawn on
sheets of paper. There were four width (W} conditions (2 in. [5.1 em];
1 in. [2.54 em]; 0.5 in. [1.3 em]; 0.25 in. [0.64 em]} and four amplitude
(A) conditions (2 in. [5.1 em]; 4 in. [10.2 em]; 8 in. [20.4 em]; 12 in. [30.5
em]). There were 16 combinations of the Wand A specifications. Paper
and pencil techniques were used in order to obtain a permanent record
of the scatter of dots in the region of the targets. Timing of the
movements was manually controlled on the initiation of the first move-
ment and on the completion of 40 taps.
192
 · Programming Precision

The subject sat at a table with his or her preferred hand midway be-
tween the targets which were placed so as to be most comfortable for
the subject. No part of the subject's body was permitted to touch the
table.
Procedure. There were two conditions:
(1) The Vision-Blind condition, in which the subjects first performed
with vision and then without vision (blind) on each of the 16 A/W
target conditions.
(2) The Blind-Vision condition, in which the subjects performed first
without vision and then repeated the same condition with vision for
each of the 16 A/W target conditions.
On the visual trials, subjects were instructed to tap alternatively be-
tween the targets as quickly and as accurately as possible, permitting
only about 10% of the dots to fall outside the target area.
With the blind trials, the subjects were instructed to shut their eyes
after the first tap. Eye closure was monitored by the experimenter
throughout the trial. Again, subjects were instructed to tap as quickly
and accurately as possible.
On each of the 16 A/W conditions, subjects were given 2 practice
trials, followed by 2 test trials. The order of presentation of the 16 target
conditions were randomly assigned to the subjects.
Downloaded by [UZH Hauptbibliothek / Zentralbibliothek Zürich] at 08:36 01 July 2014

Results and Discussion

For the data analyses, the effective target width (W) was calculated
from the scatter of hits, using the range that included 90% of the dots.
Amplitude (A) was then measured using the new values for W. The In-
dex of Difficulty was calculated in accordance with Welford's (1968)
modification of Fitts' (1954) original equation:
'
A + .SW
w
The relationship between Movement Time and Index of Difficulty for
the visual and blind conditions is summarized in Figure 1.
Both the intercept and slopes of the two functions are clearly dif-
ferent, there being a significant blind/visual condition x Index of Diffi-
culty interaction, F (14, 154) = 38.22, p < .01, and a significant
blind/visual main effect, F (1, 11) = 99.21, p < .01. There is a significant
difference for the Index of Difficulty value between the visual and blind
conditions for the corresponding target dimensions, F (1, 11) = 388.69,
p < .01, and also a significant interaction, F (14, 154) = 93.98, p < .01.
Post hoc analyses show these differences to be most significant on the
small target and large amplitude conditions. This has the effect of "com-
pressing" the upper limit of the range of Index of Difficulty values under
the blind condition (blind range = 1.7 to 3.05; visual range = 1.8 to
5.6). Indeed, the effect of the compression of the Index of Difficulty
values in blind tapping must be a consequence of changes in the actual
target width (W) and/or actual target distance (A) used by the subject.
193
 Denis Glencross and Nicholas Barrett

Further analysis showed that, under the blind conditions, the subjt=ct
virtually "ignores" the target width constraints, and for a given
amplitude, the subject uses a standard or constant target width irrespec-
tive of the target width actually specified. As the target amplitude (A) in-
creases from 2 in. (5.1 em) to 12 in. (30.5 em), the actual target wi'dth
also increases (see Figure 2), This difference is significant (p < .01 ).

Visual

<n
(.)
Ql
Blind
<n
E
-::!:
1-
-w 4
::!:
-
Downloaded by [UZH Hauptbibliothek / Zentralbibliothek Zürich] at 08:36 01 July 2014

I-
I-
zw
::!:
w
>
0
::!:

0 1 2 4 5 6 7
INDEX OF OIFFICUL TV (J.D.)

Fig. 1-Movement time for each Index of Difficulty for the blind and visual conditions.

However, as shown in Figure 3, the actual amplitude (A) closely

follows the specified amplitude with a slight tendency to undershoot at
the upper range for the 12 in. (30.5 em) amplitude conditions, There
was no difference in actual amplitude with changes in target width (W)
conditions (p > .05).
Post hoc tests showed that for each target condition (A/W), the MT for
the blind condition was significantly faster than that for the visual condi-
tion except for the 2 in. (5.1 em) target widths (see Table 1).
Table 1 shows clearly that, for the blind condition, MT remains con-
stant as a function of effective target width. Further, as target width
194
 Programming Precision

4.5 '
.&--& A= 2 ins (5.1 em)
6--tJ. -A= 4 ins (10.2 em}
0--o A = 8 ins (20.4 em)
e--e A= 12 ins (30.5 em)
4.0

~--
/ --- --·------- - ------
.
•

.
., 3.5
....... /

E
u

I ~ ----~
b
-~
3.0 -----------~--r-------- .
,_
w

"a: - --- - ---

.- ~
-'
<
2.5 .

1' -- -&-- -- - -- --- ---I;

,_
:::l
0
c( 2.0
.'
_.. _____ _----- ---- -
.. ~--__._--
-- --
1.5

0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 45 5.0

SPECIFIED TARGET WIDTH (W) ems
.

Fig. 2-Actual and specified target widths for tire various amplitude conditions for blind
Downloaded by [UZH Hauptbibliothek / Zentralbibliothek Zürich] at 08:36 01 July 2014

trials. ··
•

decreases, the error does not also decrease because the subjects do not
slow the speed of movement.
When tapping blind, it seems as though the subject is only able to pro-
gram amplitude or distance (perhaps as force x time) from the visual in-
formation available prior to the start of tapping. These programmed
details can be reasonably accurately maintained throughout the
repetitive effort. However, target width or precision apparently can-
not be programmed, for if it could be, the subject would specify a more
accurate movement for a narrower target. This does not occur. That is,
actual target width (or precision) cannot be accurately related to the
specified target width. Instead, a larger target is specified, and this larger
target is maintained for a given amplitude of movement.
Since an increase in amplitude involves an increase in the forces
specified, the changes in target width possibly reflect the changes in the
forces generated. The role of visual feedback would thus be to constrain
the variability of the response resulting from the forces generated for a
specified amplitude.
The subject makes a further adjustment when tapping blind. Except
for the very widest target (2 in./5.1 em), the subject taps significantly
faster under the blind conditions. These results indicate that for a given
amplitude at a given movement time (i.e., at a specified magnitude of
'
force being attempted), subjects will perform at the same level or
'
magnitude of error when tapping blind, irrespective of the specified •
•'

target width. This is consistent with the Schmidt et al. (1978) formula-
tion. The present findings indicate that one program is specified (one
195
 Denis Glencross and Nicholas Barrett

amplitude parameter) for each of the amplitude conditions. The target

precision value is a result of the forces generated due to movement
speed and amplitude.
Why did the visual conditions take longer than the corresponding
blind conditions? For the same A/W specification, that is, the MT for the
visual trials was significantly longer than MT for the blind trials. Further,
this effect was more obvious for the small target conditions. What these
findings suggest is that when visual feedback is available, it is used for
control and guidance (particularly to smaller targets), resulting in an
overall deceleration of the movement to achieve the desired accuracy.
This is possible for all cases where the movement time was greater than
250 msec, thus allowing time for visual feedback to be used (Keele &
Posner, 1968).
Finally, for the blind condition, when effective width (W.) is plotted
against Amplitude/Movement Time, a linear relationship is found, pro-
viding some support for the finding of Schmidt, Zelaznik, and Frank
(1978) that W = K AIMT, in the absence of visual feedback (see Figure 4).

' .

35 .._ __ .._ W = .25 ins (.635 em)

c, __ c, W = .50 ins (1.27 em)
Downloaded by [UZH Hauptbibliothek / Zentralbibliothek Zürich] at 08:36 01 July 2014

o--0 W = 1 in (2.54 em)

•-• W = 2 ins (5.1 em)

U>
E
0
-
0
0

•
-
<(

UJ
0

0
::J
1-
-
..J
0..
:2
<(
1-
UJ
Cl
a: 1
<(
1-
..J
<(
::J
1-
u
<(

0 5. 10 15 25 30 35
1
SPECIFIED TARGET AMPLITUDE (A) ems
. .

Fig. 3-Actual and specified target amplitudes for the various width conditions for blind
trials.

196
 TABlE 1
Means and S.D. for Index of Difficulty (ID) and Movement Times (MT) for Visual (V) and Blind (B) Condition in Experiment 1

TARGET TARGET WIDTH (W)

AMPLITUDE (A)
.25 in. (.635 em) .50 in. (1.27 em) 1 in. (2.54 em) 2 in. (5.1 em)

v B v B v B v B

3.09 1.98 2.19 2.00 1.95 1.85 1.84 1.69

ID
2 in. ±0.00" ±0.36 ±0.00 ±0.36 ±0.39 ± 0.31 ±0.41 ± .35
(5.1 em)
MT 406.6 274.6 319.8 243.4 265.5 248.5 233.5 226.8
(msec) ± 71.70 ±66.34 ±62.36 ± 52.58 ±58.96 + 51.57 +41.39 ±45.61
Downloaded by [UZH Hauptbibliothek / Zentralbibliothek Zürich] at 08:36 01 July 2014

4.04 2.38 3.05 2.29 2.50 2.18 2.43 2.27

ID
4 in. ±0.00 ±0.58 ±0.19 ±0.45 ±0.33 ±0.40 ±0.28 ±0.36
(I 0.2 em)
MT 488.8 273.7 377.4 269.8 326.9 257.3 270.12 242.7
(msec) ±64.38 ± 66.63 ± 64.54 ±60.74 ±65.85 ± 59.16 ±50.30 +55.02
.
•
5.02. 2.94 4.04 2.78 3.36 2.87 3.03 2.66
ID
8 in. ±0.00 ±0.39 ±0.00 ±0.37 ±0.40 ±0.33 ±0.45 ±0.34
(20.4 em)
MT 598.7 304.1 476.9 290.1 394.4 309.7 323.8 299.3
(msec) ± 91.14 ± 75.40 +56.43 ±62.43 ±81.49 ± 79.90 ± 54.88 ± 108.29 -o
~

0
5.60 3.05 4.62 2.94 3.72 3.00 3.30 2.90 00
ID ~

12 in. ±0.00 ±0.46 ±0.00 ±0.33 ±0.67 ±0.41 ±0.30 ±0.29 "'3
(30.5 em) 3
MT 686.7 336.5 537.5 304.6 430.3 312 379 323
-·
:::l
00
(msec) ±110.90 ±89.48 ±47.86 ± 75.43 ±55.95 ± 76.40 ± 75.00 ± 72.00 -o
~-·
~

"S.D. of 0.00 indicates that subjects replicated the specific Amplitude (A) and Width (W) conditions exactly. "'-·
0
"'
" :::l
 Denis Glencross and Nicholas Barrett

However, Schmidt et al. (1978) predictt?d that the function relating

AIMT to w. should pass through the origin. This is inconsistent with
Figure 4.

General Discussion and Conclusion

It has been proposed that movement time increases with increases in
the distance to be moved and the precision required for the outcome
because there are limits to the accuracy of the translation of the task re-
quirements to the actual movement. Beggs and Howarth (1972) refer to
a constant angular error in aiming (00) which, presumably, reflects the
accuracy of translation. Schmidt et al. (1978) have proposed that this ac-
curacy of translation may be related to the force attempted by the sub-
ject. The present paper examined movements to targets and transla-
tional accuracy, both in the presence and in the absence of visual feed-
back. It is proposed that making rapid movements to targets, without
access to visual information, necessitates the subject to program the ac-
tion. The spatial requirements must be translated into a precise move-
ment pattern which cannot be modified during the ongoing movement
by visual feedback.
In general, the results of the present study point to a limit in the
Downloaded by [UZH Hauptbibliothek / Zentralbibliothek Zürich] at 08:36 01 July 2014

translational accuracy of perception to action. Though amplitude may

be reproduced accurately, width is constant for a set amplitude. There
was, furthermore, a suggestion that a subject may maintain translational
accuracy by responding more rapidly in the repetitive tapping situa-
tion rather than slowing the individual movements in the blind condi-
tion to increase accuracy, the subject tries to tap faster, thereby com-
pleting more individual movements within a set time.
For blind movements, as the amplitude of the movement increases,
the effective target width increases independently of the specified target

5 A (em) •

*0 10.2
5.1

* 20.4
4- * 30.5

·~

3: 3-
•
l'o
*
•

0 ---.---.--------r---------.----------,
o 1~ ~ 10 ~ so w ~ ao qo 1do
AMPLITUDE/MOVEMENT TIME (em/sec)

Fig. 4-The relation between effective target width (W.) and the ratio of the movement
amplitude (A) and the movement time (MT), for blind trials only.
'

198
 Programming Precision

width. The precision of the movement is dictated instead by the ampli-

tude of the movement. It is possible that when the required precision
might exceed the inherent translation accuracy of the system, then
recourse is made to visual feedback.
Overall, these results suggest a distinction between distance covering
processes and corrective processes. This distinction was first suggested
by Woodworth (1899) and later by Welford, Norris, and Shock (1969).
They calculated separately the slopes relating movement time to dis-
tance and to precision, reporting results which fit the data better.
Distance covering processes contributed to the magnitude of move-
ment time in the blind condition. Movement time was related to ampli-
tude. The difference in results obtained in the visual condition over the
blind condition, however, may be attributed to corrective processes.
The tapping speed for blind and visual trials was similar for the large
width conditions, regardless of the amplitude. For small target widths,
visual movements were slower than the blind movements. The relative
roles, however, of programming and proprioception in the distance
covering processes are not known. Precision, however, appears to be
specifically the domain of visual feedback. For a set amplitude, the ex-
ecutive appears to generate a movement of a set variability. Visual feed-
back helps reduce this variability. The distance, in blind movements,
Downloaded by [UZH Hauptbibliothek / Zentralbibliothek Zürich] at 08:36 01 July 2014

can be specified quite accurately, and consonant with this, there is a

magnitude of force generated and a proportional movement endpoint
variability. As such, the results may be expected to behave in a manner
predicted by Schmidt et al. (1978). Rather than reducing force (speed) to
gain precision, however, subjects chose to minimize the temporal inter-
response space when tapping repetitively. This suggests that subjects
adopt strategies in the discrete movement paradigm (Schmidt et al.,
1978) different from the repetitive tapping paradigm. This is an impor-
tant observation deserving more research.
Though, when tapping blind, subjects did not reduce their speed to
further reduce movement endpoint variability, Figure 4 nevertheless
suggests that effective target width under blind conditions is linearly
related to the ratio of amplitude and movement time. This is a restate-
ment of Schmidt et al.'s (1978) claim that impulse variability, and as a
result W., is directly proportional to movement amplitude and inversely
proportional to the movement time. The function relating W, to AIMT,
however, did not pass through the origin. This departure from propor-
tionality was also observed by Schmidt et al. (1978) and poses problems
for their model, needing further elucidation.

REFERENCES
Beggs, W. D. A., & Howarth, C. I. The movement of the hand towards a target. Quarterly
journal of Experimental Psychology, 1972, 24, 448-453.
Crossman, E. F. R. W., & Goodeve, P. ). Feedback control of hand-movement and Fitts'
Law. Proceedings of the Experimental Society, Oxford, 1963.
Fitts, P. M. The information capacity of the human motor system in controlling the
amplitude of movement. Journal of Experimental Psychology, 1954, 47, 381-391.

199

I •
."
'• .
 Denis Glencross and Nicholas Barrett

Howarth, C. 1., Beggs, W. D. A., & Bowden,). M. The relationship between speed and ac-
curacy of movement aimed at a target. Acta Psycho/ogica, 1971, 35, 207-218.
Keele, S. W. Movement control in skilled motor performance. Psychological Bulletin,
1968,70,387-403. i

Keele, S. W. Behavioral analysis of motor control. In V. B. Brooks (Ed.), Handbook of

physiology: Motor control (vol. Ill). Washington, D.C.: American Psychological Society,
1980.
Keele, S. W., & Posner, M. I. Processing of visual feedback in rapid movements. journal of
Experimental Psychology, 1968, 77, 155- i 58.
. '
langolf, G. D., Chaffin, D. B., & Foulke, J, A. An investigation of Fitts' law using a wide
range of movement amplitudes. Journal of Motor Behavior, 1976, 8, 113-128.
Megaw, E. D. Possible modification to a rapid ongoing programmed manual response.
Brain Research, 1974, 77, 425-441.
Schmidt, R. A., Zelaznik, H. N., & Frank, J. S. Sources of inaccuracy in rapid movement. In
G. E. Stelmach (Ed.), Information processing in motor control and learning. New York:
Academic Press, 1978.
Vince,. M. A. Corrective movements in a pursuit task. Quarterly Journal of Experimental
Psychology, 1948, 1, 85-103.
Welford, A. T. Fundamentals of skill. London: Methuen, 1968.
Welford, A. T., Norris, A. H., & Shock, N. W. Speed and accuracy of movement and their
changes with age. In W. G. Kosier (Ed.), Attention and performance II, 3-15. Amsterdam:
North Holland Publishing Co., 1969.
Woodworth, R. S. The accuracy of voluntary movement. Psychological Review, 1899, 73,
1-114.

Submitted April 17, 7981

Downloaded by [UZH Hauptbibliothek / Zentralbibliothek Zürich] at 08:36 01 July 2014

Final revision submitted December, 1982

200