Speed Accuracy Trade-Off 1
Running Head: SPEED AND ACCURACY TRADE-OFF IN MANUAL AIMING TASKS
AND PURSUIT ACCURACY
Speed and Accuracy Trade-Off in Manual Aiming Tasks and Pursuit Accuracy
A laboratory report
presented to Tara Artibello
in HKIN 215
Intro to Motor Learning
Department of Human Kinetics
St. Francis Xavier University
December 4, 2009
Speed Accuracy Trade-Off 2
In 1954 Paul M. Fitts found target acquisition time to have a linear relationship with the
information load or difficulty of a movement. This was discovered through an experiment now
dubbed Fitts’ pointing paradigm (Guiard & Beaudouin-Lafon, 2004). This was a variance of
Fitts’ initial prediction that movement time would be directly proportional to the information
load of a movement (Guiard & Beaudouin-Lafon, 2004). Since then, this widely applicable
finding has been implemented in a multitude of situations. Studies such as that of Seya and Mori
(2007) in which the trade-off between motor time and pursuit accuracy was studied have greatly
expanded upon Fitts’ law. This experiment was designed to test Fitts’ findings in a traditional
manner by way of a manual aiming task with variable complexity or difficulty of movement.
There should be an increase in motor time as the difficulty of the movement increases. Thus an
increase in motor time will correspond with a decrease in target size or an increase in the
distance between targets.
The participant in this study was an undergraduate student at St. Francis Xavier
University who was enrolled in HKIN 215. This single participant took part in all four conditions
of the study.
Materials used in this experiment were target sheets, one for each condition, a pen and a
The participant performed three trials for each condition, in random order. Each condition
had a varying degree of difficulty depending upon the distance between the targets and the width
Speed Accuracy Trade-Off 3
of the targets. Each trial had a duration of ten seconds, which was timed by the experimenter.
The experimenter signaled the beginning and the end of each trial with the words “ready - go”
and “stop” respectively. During each trial the subject attempted to tap the pen within the two
targets on the target sheet alternately as quickly as possible. Between each trial there was a rest
period equivalent to the time it took to count the taps within each of the two targets for that
particular trial. If the number of errors reached or exceeded five percent, the trial was repeated.
An error was considered to be a pen mark outside the designated target area on the target sheet.
In condition A the distance between targets on the target sheet was two centimeters and
the width of each of the two targets was two centimeters. In condition B the distance between
targets was four centimeters and the width of each target was two centimeters. In condition C the
distance between targets was eight centimeters and the width of each target was one centimeter.
In condition C the distance between targets was sixteen centimeters and the width of each target
was two centimeters. The independent variables in the study were both the distance between
targets and the width of each target. The dependant variable in the study was the cumulative
number of pen marks within the two targets.
For condition A the number of taps per second was 4.1, the mean motor time for this
condition was 243.9. The number of taps per second for condition B was 3.5 and the mean motor
time was 285.7. The number of taps per second for condition C was 2.2 and the mean motor time
for this condition was 454.5. For condition D the number of taps per second was 2.5 and the
mean motor time was 400. For further data summary see Table 1. Mean motor time was
Speed Accuracy Trade-Off 4
calculated by taking the inverse of the average number of taps per second then, multiplying this
number by 1000 to get the average motor time in milliseconds.
Using target sheets to measure the number of taps per second and further more calculate
mean motor time, this study was able to evaluate the speed and accuracy trade-off of a manual
aiming task as described by Paul Fitts. The number of taps per second decreased as the distance
between targets increased and as the width of the targets decreased. Mean motor time increased
as distance between targets increased and as the width of the targets decreased. These results
confirm a trade-off between speed and accuracy in a manual aiming task. The data showed a
decrease in motor time as the difficulty or complexity of the task increased, either with an
increase in the distance between targets or a decrease in the width of the targets.
Similar results were found in a study done by Seya and Mori (2007) in which the reaction
time to a visual target was measured during smooth pursuit of a moving fixation stimulus. Seya
and Mori (2007) found that participants displayed a decrease in pursuit gains as the velocity of
the fixation stimulus increased. A pursuit gain in this study was measured as the ratio of eye
velocity to the fixation stimulus velocity. From the decrease in pursuit gains it can be deduced
that participants may have reduced pursuit accuracy in order to achieve faster reaction times
(Seya & Mori, 2007). In other words there is a trade-off between pursuit accuracy and reaction
times (Seya & Mori, 2007). Murata and Iwase (2001) extended Fitts’ law further and attempted
to apply it to a three-dimensional pointing task in which the difficulty of the movement was
influenced by target size, distance to the target and direction to the target. In this study the task of
the subject was to point to a target as directed by the experimenter, with their right index finger.
There were significantly more variable movement times in this study as there was a significant
Speed Accuracy Trade-Off 5
effect of movement direction on movement time (Murata & Iwase, 2001). Although the
conventional Fitt’s law was modified to increase the predictive power on motor time by the
addition of variables, specifically a directional term, the performance model is still applicable to
a certain degree to three-dimensional pointing tasks (Murata & Iwase, 2001).
Tresilian, Plooy and Marinovic (2008) also confirmed Fitts’ prediction of a speed-
accuracy trade-off in a more complex movement situation by testing the belief that movement
time is typically greater when there is a greater accuracy requirement in an interceptive aiming
task. In this experiment subjects were required to move a hand held manipulandum along a
vertical track, into the path of a moving target. Thus in this case movement difficulty was
dependent upon a temporal accuracy requirement and a spatial accuracy requirement. The typical
response to a greater requirement of spatial accuracy was found to be longer duration
movements. This response was also found when subjects were required to hit smaller targets
(Tresilian et al., 2008). These findings support those of this study and support the notion that
longer motor times are a typical response to greater accuracy requirements.
Speed Accuracy Trade-Off 6
Guiard, Y., & Beaudouin-Lafon, M. (2004). Fitt’s law 50 years later: applications and
contributions from human-computer interaction. International Journal of Human-
Computer Studies, 61, 747-750.
Murata, A., & Iwase, H. (2001). Extending Fitt’s law to a three-dimensional pointing task.
Human Movement Science, 20, 791-805.
Seya, Y., & Mori, S. (2007). Tradeoff Between Response Speed and Pursuit Accuracy. Motor
Control, 11, 109-118.
Tresilian, J., Plooy, A., & Marinovic, W. (2009). Manual interception of moving targets in two
dimensions: Performance and space-time accuracy. Brain Research, 1250, 202-217.
Speed Accuracy Trade-Off 7
Condition D(cm)/W(cm) Median Taps Number of Taps Mean Motor
Per Second Time (msec)
A 2/2 41 4.1 243.9
B 4/2 35 3.5 285.7
C 8/1 22 2.2 454.5
D 16/2 25 2.5 400.0