Multi-aspect Evaluation Method for Digital Pointing Devices

Nils B

uscher, Daniel Gis, Sebastian Stieber and Christian Haubelt

Institute of Applied Microelectronics and Computer Engineering, University of Rostock, Rostock, Germany

Keywords:

Evaluation, Pointing Device, User Experience, Inertial Sensors.

Abstract:

For decades the computer mouse has been used as the most common input device for laptops and computers

alike. However for speeches a presentation remote with a laser pointer was used because they allowed the pre-

senter more freedom. With the emergence of small and lightweight inertial sensors, a new type of presentation

remotes becomes popular. These remotes use inertial sensors to move a digital pointer allowing presenters to

show things on more than one screen or use enhancement methods like highlighting a region. Using inertial

sensors however proves to be a difﬁcult task and can lead to problems with the usability of such devices.

When developing such systems, the designer faces the problem that no method for quantifying the usability of

pointing devices based on inertial sensors is available. In the paper at hand, we propose an evaluation method

consisting of three different tests to assess the manageability, speed and precision of digital pointing devices

for a measurable comparison. Additionally, we conducted an evaluation to show that our tests reﬂect the sub-

jective assessment from the users. Our quantitative test results showed a strong correlation to the qualitative

subjective assessment from the users.

1 INTRODUCTION

The usage of Digital Pointing Devices (DPD) for pre-

sentations has many advantages compared to tradi-

tional devices like e.g. laser pointer. Especially when

a presentation is shown on multiple screens simulta-

neously, the usage of a digital pointer has a huge ben-

eﬁt because the presenter can point on all screens at

the same time and does not have to turn around to

look at the screen to point at it. In addition it allows

the presenter to utilize a number of digital enhance-

ment methods during the presentation like highlight-

ing a certain region of the presentation slides or do-

ing annotations during the presentation. The usage of

a DPD however is limited by the accuracy in sens-

ing its movement. Additionally, the sensitivity and

movement proﬁle of the device have an inﬂuence on

its usability. Ideally the usage of a digital pointing

device should be as intuitive as using a laser pointer.

The usability of such a DPD is the main objective in

its design and needs a lot of attention and ﬁne tuning

during its development to be suitable for a wide user

base.

In (MacKenzie and Jusoh, 2001) it is shown that

these devices are often far away from the usability of

a standard computer mouse and improvements need

a lot of attention to bring them to the same level of

usability. It is, however, not easy to determine the us-

ability of such a device as there are multiple aspects

that have to be considered at the same time. First of

all the subjective assessment of the user about the per-

formance is of great importance. Second, the DPD

should allow the user to be able to precisely point

at a target on the screen and also be able to quickly

move between multiple targets to allow for a ﬂuent

presentation. However, no comprehensible and accu-

rate evaluation method to assess the performance and

usability of DPDs is available today.

This paper, proposes the multi-aspect evaluation

method to asses and compare different DPD imple-

mentations. Our evaluation method consists of three

separate tests that evaluate different aspects of the

device: The ﬁrst test evaluates the responsiveness,

which we see as the ability to move the pointer di-

rectly between two objects on the screen. The second

test evaluates how fast a user can move the pointer be-

tween objects on the screen. The third test evaluates

how precise the pointer can be moved along lines or

shapes. To evaluate our testing approaches, we con-

ducted tests for four pointing devices with multiple

participants and compared the measured results from

the tests against each other and against the subjective

assessment from the testers.

The remainder of paper is structured as follows:

In Section 2, this paper discusses other publications

dealing with the assessment of the usability of digital

128

Büscher, N., Gis, D., Stieber, S. and Haubelt, C.

Multi-aspect Evaluation Method for Digital Pointing Devices.

DOI: 10.5220/0008355701280135

In Proceedings of the 9th International Conference on Pervasive and Embedded Computing and Communication Systems (PECCS 2019), pages 128-135

ISBN: 978-989-758-385-8

pointing devices in general. Afterwards in Section 3,

we explain which criteria were used for the evalua-

tion methods and describe the three tests which were

developed to cover said criteria. In the following Sec-

tion 4, an evaluation of the proposed methods is done

using 160 measurements from multiple participants.

Finally, Section 5 concludes this paper and gives a

short outlook for future development.

2 RELATED WORK

In the past there have been many studies about the

usability of digital Human Interface Devices (HIDs).

The performance evaluation shown in (Batra et al.,

1998) examines the usability of different built-in in-

put devices for laptops, namely trackballs, stick points

and touch pads. Said research tested three common

tasks performed on a computer normally performed

with a computer mouse and used the needed time,

click errors and click accuracy. An assessment of the

ergonomics of the used input devices was also con-

ducted. The results however were meant to compare

the usability of the devices and allow to compare sub-

jective user experience with the measured results.

The International Organization for Standardiza-

tion (ISO) released a norm to assess the usability of

digital input device (ISO, 2007). However, as re-

vealed in (Douglas et al., 1999) this standard does not

allow a direct correlation between the measured per-

formance and the user experience. To see how well

such a pointing device will be accepted by potential

users our approach seeks to create a direct correlation

between measured performance and user experience.

Another work that seeks to examine the perfor-

mance of input devices is (MacKenzie et al., 2001),

which also derives multiple measurable aspects of the

devices to evaluate their performance. The focus of

this work however is on devices used for common of-

ﬁce tasks on a computer and measures how fast and

efﬁcient certain tasks can be done. It does not con-

sider the assessment of the users on how usable and

comfortable the devices are.

3 TEST METHODS

For the evaluation of digital pointing devices, the ﬁrst

important step was to identify which properties can

be used for an objective assessment and comparison.

Three major properties were found:

Manageability. The manageability describes how

well a device can be handled by the user to ex-

ecute simple and common actions. For example

how easy it is for a user to move from one place

on the screen to another.

Responsiveness & Speed. The responsiveness and

speed is a factor for how fast a user is able to

conduct simple and common actions. In particular

the responsiveness describes how fast or sluggish

the device reacts to the user input and the speed

describes how fast the user is able to move the

pointer.

Precision & Stability. The precision and stability

properties describe how well a user can conduct

more complex actions. This can be for example

highlighting a certain small object on the screen,

write a note or move along a complex shape.

For the evaluation of the digital pointing devices,

three tests were designed, where each evaluates one

of the three properties mentioned above. The ﬁrst test,

called the ’Pointing Test’ evaluates the manageability.

In particular, it assesses how well a user can move the

pointer from one place of the screen to another place.

The second test evaluates the speed and responsive-

ness. It is measured how fast a user can move the

pointer between different places on the screen. The

last test evaluates the precision and stability of the

pointing device. It measures how well a user can trace

the shape of different objects with the help of the dig-

ital pointing device.

3.1 Pointing Test

The Pointing Test was designed to evaluate manage-

ability of the pointing device. It tests how well the

user is able to move directly between points on the

screen. At the beginning, the user has to move the

pointer to the ﬁrst point on the screen. When the ﬁrst

point is reached, a second point appears after a short

time to which the user has to move in a direct path.

After reaching that point, the next point becomes vis-

ible and the user has to move to the new point. The

user has to stay with the pointer in the current point

for at least two seconds before the new point appears

to evaluate that it is possible to reach a target point

and also stay there. This procedure is repeated multi-

ple times.

In the GUI the current point is drawn in blue while

the point to move to is drawn in red. The test ﬁnishes

after the user has traversed a number of points that

was deﬁned prior to the start of the test.

The score for the evaluation in the Pointing Test is

calculated via the relation between optimal (minimal)

distance between the points and the actually moved

distance in pixels. A shorter distance traveled with

the pointer results in a higher score. Equation 1 shows

how the score for the traveled distance is calculated.

Multi-aspect Evaluation Method for Digital Pointing Devices

129

Current Point

Next Point

Pointer

Traversed Point

Ideal Path

Taken Path

Overshoot

Figure 1: Top: Interface of Pointing Test and Speed Test.

Bottom: Results from Pointing Test and Speed Test.

score

dist

·dist

min

+(1−α

dist

)·(dist

min

−dist

traveled

)

dist

∗dist

min

(1)

The score score

dist

for the distance is calculated as

the ratio between the minimal distance dist

min

and the

traveled distance dist

traveled

. The difference between

minimal distance and traveled distance is weighted by

dist

to allow for a suitable scaling of the score. With-

out the weighting the score would reach zero when

the traveled distance is at least twice as long as the

minimal distance. This scenario however is very un-

likely and initial tests with multiple devices showed a

way shorter traveled distance even for the worst de-

vices. Based on the initial measurements with the test

we chose an α

dist

of 0.25, so the score will reach zero

when the traveled distance is 33.

3% higher than the

minimal distance. Said factor was chosen empirically

to allow for a better scaling of the score for the trav-

eled distance.

In addition to the traveled distance, also an over-

shoot distance is used for the score to respect the fact

that the user might move past the target point while

trying to reach it. An overshoot can be seen at the bot-

tom of Figure 1. The red line is the traveled path from

the pointer after the user reached the target point but

did not start to move to the next point. This so-called

overshoot might happen if the device reacts too slow

to the user input or is not well manageable. The score

for the overshoot is calculated in Equation 2.

score

overshoot

dist

min

− (α

overshoot

· dist

overshoot

)

dist

min

(2)

A higher overshoot will result in a lower score as it

shows that the device does not allow the user to stay

in the reached location. The overshoot is calculated

by the ratio between the minimal distance that can

be traveled dist

min

and the traveled distance from the

overshoot dist

overshoot

weighted by α

overshoot

. Based

on initial tests, the factor α

overshoot

is also set to 0.25

to allow for a better scaling of the range from the score

because it is very unlikely that the overshoot distance

ever reaches the actual distance the user has to travel.

We use the minimal distance for all scores as a

factor to allow for a consistent evaluation across mul-

tiple screens with different resolutions. Using a ﬁxed

value for the evaluation would result in different re-

sults depending on the screen size and resolution as

shown in (Oehl et al., 2007). Additionally the tests

were conducted on a huge screen to reﬂect the com-

mon use case. The overall score score

pointing

for the

Pointing Test is calculated by Equation 3.

score

pointing

= score

dist

· score

overshoot

(3)

The values for score

dist

and score

overshoot

are clamped

to not be below 0 before the calculation. A score

higher than 1 should not be possible.

3.2 Speed Test

The composition of the Speed Test is similar to the

composition of the Pointing Test. Again the user has

to move to the ﬁrst point and to all subsequent points

like in the Pointing Test. However, the Speed Test

evaluates how fast a user can move between multiple

points using the digital pointing device. Therefore the

test participants do not have to move in a direct line

between the points but try to reach as many points

possible in a given time frame.

The test only considers the time the user spends

moving between the points on the screen and does not

count the time the user stays at a point. This should

prevent that the reaction time from the user inﬂuences

the test result when the points are placed at random

locations on the screen.

The score for the Speed Test is calculated by the

relation between available time and reached points.

The Equation for the score calculation can be seen in

Equation 4.

score

speed

num

points

· β · f

distance

time

(4)

The time is the overall time for the test in millisec-

onds. The factor β is used to weight the number

of points against the time, so that it can be set how

many points have to be reached per second to get the

PECCS 2019 - 9th International Conference on Pervasive and Embedded Computing and Communication Systems

130

highest score. Based on initial test we chose a fac-

tor of β = 500 for two points per second. The factor

distance

plays an important role for the repeatability

and accuracy of the speed test. The speed test is de-

signed that the points are placed at random locations

on the screen to prevent a habituation effect of the

tester. However, the random placement of the points

causes the overall distance the tester has to travel to

vary between each test. For example, in the ﬁrst test

all points are placed relatively close to each other and

the tester manages to reach eight points. In the second

test, the points are very far away from each other and

the tester only reached four points and gets a worse

score although the traveled distance might be actually

bigger than in the ﬁrst test. The f

distance

is a weighting

factor used to incorporate the distance that is traveled

and is calculated by Equation 5.

distance

dist

traveled

num

points

· dist

average

(5)

The dist

average

is the average distance between all pos-

sible points on the screen and can be calculated at the

beginning of the test. Using this factor the score is in-

creased when the average distance traveled dist

traveled

between the points is longer than the expected average

distance and decreased when the traveled distance is

shorter. We chose to use the square root of the ratio

from traveled and expected distance to consider the

fact that a higher distance between points also allows

the user to move the pointer faster. The number of

reached points is num

points

3.3 Precision Test

The purpose of the Precision Test is to evaluate the

precision and stability of the digital pointing device.

A precise movement is important to be able to point

at small objects or interact with the content on the

screen, like drawing a shape or highlighting an object.

For the Precision Test the tester has to move the

pointer along the edge of multiple shapes to trace

them. The difference between the drawn shape from

the pointer and the real shape is used to calculate the

score for the pointing device. The interface of the Pre-

cision Test can be seen in Figure 2.

For the Precision Test three shapes have to be

traced with the pointer: The red and blue points are

used as information for the user to show the progress

and to ensure that the user really traces the shape.

Square. With the square it can be tested how well the

pointing device allows a user to move precisely in

vertical and horizontal directions.

Pointer

Reached

Point

Points

still to

reach

Taken path

Shape to move along

Movement

errror

Figure 2: Interface of the Precision Test during the test.

Bottom: Interface with the results. The red area is the move-

ment error.

Triangle. The triangle is used to evaluate how well

the pointing device allows a user to move diago-

nally between elements on the screen or draw di-

agonal lines.

Circle. With the circle the test can evaluate how

well a user can move along curved edges or draw

curved shapes.

With the three shapes different aspects and character-

istics of the digital pointing device can be evaluated.

The ﬁrst part for the score of the Precision Test is cal-

culated with the Equations 6 and 7.

dist

avg

∑

n=0

distanceShape

(6)

Here distanceShape

is the shortest distance between

the n

pointer position and the edge of the shape. N is

the number of sample points captured during the test.

score

shape

= 1 −

γ · dist

avg

shapeCircum f erence

(7)

The average distance between the shape and the tra-

versed path it multiplied by the factor of γ to allow for

a better scaling of the score score

shape

and was deter-

mined empirically to be 25. The circumference of the

shapes is shapeCircum f erence.

In addition to the precision determined by the dis-

tance between the shape and the traversed path, it is

also important to evaluate the stability of the pointer.

Therefore a second score was added to the Precision

Test which uses the sum of all angles of the traversed

path to compare it with the optimal angle. For all

shapes the optimal angle for a movement would be

Multi-aspect Evaluation Method for Digital Pointing Devices

131

360°. However an unstable pointer will have a much

higher angle at the end because it moves up and down

or left and right while being moved along the shape.

This effect is illustrated in Figure 3 which shows two

paths for a stable and for an unstable pointer including

the angle change between each movement.

Taken path

Optimal path

Δ41°

Δ74°

Δ90°

Δ31°

Figure 3: Traversed path of a stable pointer (blue) and ex-

emplary path of an unstable pointer (red).

When summing up the delta angles from the blue

path, the overall angle change of the path is 90°. For

the red path the angle changes sum up to 146°. Thus

a higher angle change indicates an unstable pointer.

To not add huge errors for very small movements,

where the pointer jumps between pixels adding an-

gles of up to 180°, the movement angles are only cal-

culated when the pointer moved a deﬁned number of

pixels away from the last location where the angle was

calculated.

Using the angles, the second part of the score is

calculated by Equation 8.

score

angle

= 1 −

1 −

angle

min

angle

traversed

(8)

Here angle

min

is 360 divided by the number of cal-

culated angles and angle

traversed

is the average of all

angle deltas from the measurement. The overall score

for the precision test is then calculated by Equation 9.

score

precision

= score

shape

· score

angle

(9)

The total score score

precision

of this test is the product

of the score for the shape and the score for the angles.

4 EVALUATION

For the evaluation of the proposed test methods we

conducted tests with participants of different ages and

sex on four devices. A huge diversity of the partici-

pants was important because it was shown in (Char-

ness et al., 2004) that these factors have an not negli-

gible inﬂuence on the results.

The devices used for the evaluation were:

• A computer mouse

• A commercially available pointing device

• A smartphone

• A prototype of a digital pointing device

The prototype consisted of a 1.5 x 3.0 cm sensor board

rigidly attached to a ballpoint pen. The mouse was

used with a mouse-pad on a ﬂat table, the other point-

ing devices were used like a laser-pointer pointing at

a 55” LCD monitor 3 meters away.

The computer mouse is used as a reference device

because it is the most commonly used input device for

computers (Atkinson, 2006) (Greenemeier, 2009).

To counteract a habituation effect when using the

pointing devices, the sequence at which the test par-

ticipants tested the devices was randomized. Over-

all, 160 measurements were made for all tests com-

bined. After test conduction, the participants were

asked for their subjective assessment of the pointing

devices where they had to sort the devices from best

to worst regarding their usability and handling.

4.1 Expected Results

To proof the reliability of the results from the pro-

posed evaluation methods we expected that the results

from the tests and the subjective assessment of the

users show similar tendencies. Similar results would

support the assumption that the ability to use the digi-

tal pointing device in a fast and precise manner would

lead to a positive subjective assessment from the par-

ticipants.

It was also expected that the results from the

mouse would have the best results and therefore be

the reference against which the other devices are

compared, which was already shown by (MacKenzie

et al., 2001). The mouse proofed to be one of the most

precise pointing devices used in the last decades and

nearly everyone uses a mouse on a regular basis when

controlling a computer.

4.2 Pointing Test

The results from the Pointing Test are shown in Figure

4. The Pointing Test showed that the results from the

mouse had the best overall score, directly followed by

our prototype. The commercial product is placed on

the third place. Moving the pointer with a smartphone

showed the worst results.

The results indicate that the prototype has nearly

the same performance as the mouse. The inferior be-

haviour of the smartphone was caused by a notice-

able latency and some special characteristics that are

pointed out in more detail in Section 4.4. The results

still show that the minimal score reached from the

PECCS 2019 - 9th International Conference on Pervasive and Embedded Computing and Communication Systems

132

Prototype

Commercial Smartphone

Mouse

0.5

0.6

0.7

0.8

0.9

Pointing Score

Figure 4: Score for the pointing test from the four tested

devices.

smartphone was slightly below 0.7. Hence, all tested

devices were still usable.

4.3 Speed Test

The results of the Speed Test show a similar order

for the score from the tested devices. However it is

clearly visible in Figure 5 that the results from the

mouse are a lot better than the results from the other

devices. A superiority of the mouse was expected for

this test. First of all, all testers were already famil-

iar with the usage of a computer mouse, secondly the

movement needed to move the pointer with the mouse

was different from the other tested devices which di-

rectly pointed at the screen to move the pointer. Last

but not least, the mouse was the only device directly

connected to the computer. The commercial pointer

and the smartphone communicated wirelessly with

the computer. The prototype was connected via an

additional shuttle board. Therefore the latter three

devices have a higher latency between device move-

ment and pointer movement, which has an effect on

the overall performance of said devices.

Prototype

Commercial Smartphone

Mouse

0.5

0.6

0.7

0.8

0.9

Speed Score

Figure 5: Score for the speed test from the four tested de-

vices.

Similar to the Pointing Test, the prototype showed

the second best results followed by the commercial

product. The smartphone again had the worst results.

However, the difference between the three devices is

lower than in the Pointing Test. This is caused by

the way the score is calculated to also incorporate the

score from the mouse in a reasonable manner. Ad-

ditionally the nature of the test using the number of

reached points for the score causes a lower variance

as it was possible to reach nearly the same number of

points with all three pointing devices. It is possible

that a longer duration of the test would yield better

distinguishable results.

4.4 Precision Test

The score for the Precision Test consists of three sub-

scores for the movements around the square, the trian-

gle and the circle. The overall score of the Precision

Test shows a different distribution for the usability of

the devices, than the Pointing Test and the Speed Test,

due to an anomaly in the test with the square shape.

The sub-scores and the combined score can be seen in

Figure 6.

Prototype

Commercial Smartphone

Mouse

0.5

0.6

0.7

0.8

0.9

1.1

Precision Score

Square

Triangle

Circle

Overall

Figure 6: Score for the precision test from the four tested

devices.

4.4.1 Square Shape

The scores from the evaluation with the square shape

show different results than all other conducted tests.

It can be seen in Figure 6 that the Smartphone has

the best score in this part of the test despite being the

worst for the two other shapes and the other tests. The

second best score is from the Mouse, followed by the

Prototype. The worst result was measured with the

commercial product. While the score for the smart-

phone was the worst in all other tests it shows the best

results for the Precision Test with the square shape.

This anomaly is caused by the combination of three

factors:

1. All edges of the square shape are either vertical or

horizontal.

2. To reduce the inﬂuence of noise from the sensor-

data, the application on the smartphone uses a

Multi-aspect Evaluation Method for Digital Pointing Devices

133

high threshold. All accelerations and angle-rates

below that threshold are ignored.

3. The threshold is applied independently for each

single axis on the smartphone.

When the user now moves the pointer along the

shape of the square, the movement on one axis is

low enough to be below the threshold and is ﬁltered

out causing the pointer to get ’locked’ on one of the

axis. Therefore the pointer can be moved in a perfect

straight line which makes it a lot easier for the user to

draw the square shape.

4.4.2 Triangle Shape and Circle Shape

The tests for the triangle shape and the circle shape

showed similar results to the Pointing Test and the

Speed Test. The mouse has the highest score, fol-

lowed by the prototype and the commercial device.

The results from the Smartphone are again the worst,

which is not only caused by the overall bad accuracy

of the used device. Additionally, the properties de-

scribed in 4.4.1 make it impossible for the user to

move along a diagonal line or curve because one of

the movement axis very often gets ’locked’ causing a

sawtooth distortion of the actual movement.

4.4.3 Combined Score

The combined score from the Precision Test clearly

shows the inﬂuence of the outlier measurement from

the square shape causing the score from the smart-

phone to be higher than the score of the commercial

pointing device. All other devices show a similar re-

lation compared to the Pointing Test and Speed Test.

The distribution of the results shown in Figure

7 show a relatively constant performance from the

Mouse and the commercial product. The prototype

had worse results in some cases indicating a need for

improvement. The results from the smartphone had

the highest distribution indicating that it cannot be re-

liably used for the given use-case.

Prototype

Commercial Smartphone

Mouse

0.4

0.6

0.8

Overall distribution

Figure 7: Distribution of the test results.

4.5 Overall Score

For the comparison of the test methods against the

subjective user assessment, the average score for each

device was calculated and normalized to be between

0 and 1. The resulting scores can be seen in Figure 8.

Prototype

Commercial Smartphone

Mouse

0.6

0.7

0.8

0.9

1.1

Average Score

Figure 8: Normalized average score for the tested devices.

It can be seen that the outlier from the measurement

of the Precision Test with the square shape did not

have a huge inﬂuence on the average score.

Using other values for the scaling factors α, β

and γ, the overall score will show a different scaling.

However, changing these factors will not change the

relative score between the devices. A change of α, β

and γ also changes the inﬂuences of each single test

in the overall score.

4.6 Subjective User Assessment

The subjective assessment of the usability of the

tested devices allows us to determine if our developed

evaluation methods can be used to make a qualitative

comparison of the digital pointing devices in a reli-

able way that reﬂects the experience of the users.

For the subjective assessment the user had to put

each device into an order from best to worst. The

best device receives 3 points, the second one 2 , the

third 1 and the last one 0 points. The results from the

subjective assessment from the test participants can

be seen in Figure 9.

Prototype

Commercial Smartphone

Mouse

0.2

0.4

0.6

0.8

User Score

Figure 9: Normalized subjective usability assessment.

As expected, the mouse achieved the highest score

and was rated to have the best usability from all test

PECCS 2019 - 9th International Conference on Pervasive and Embedded Computing and Communication Systems

134

participants. The qualitative rating of all evaluated

devices reﬂects the rating that was measured by the

three test methods except for the square shape from

the Precision Test due to the named factors from Sec-

tion 4.4.1.

4.7 Comparison

The results from the subjective user assessment and

the scores from the tests show a strong correlation,

which suggests that the proposed test methods can be

used to make a qualitative assessment of the usability

of a digital pointing device. Apart from one exception

all tests show the same order for the usability of the

tested devices as the subjective user assessment.

However, the results show that the subjective user

assessment has to be enhanced to allow for a quanti-

tative evaluation with the developed testing methods.

Therefore the user should not only create a ranking

for the devices but also rate them between 0 and 10

for manageability, responsiveness and precision.

5 CONCLUSION

In this paper, we presented three test methods to eval-

uate the usability of digital pointing devices whose

primary focus is on presentations. Each of the test

methods evaluates one of the three identiﬁed proper-

ties of a digital pointing device that can indicate its

usability, namely manageability, speed and precision.

Each of the tests calculates a score to assess the tested

devices. In the evaluation we showed that the assess-

ments from the proposed tests were able to reﬂect

the subjective assessment from test participants and

therefore showed that the tests are able to compare

and evaluate the devices in a qualitative way.

The formulas for the tests use many empirically

determined factors that showed to give reasonable re-

sults during initial tests and were chosen in a way to

cover the range of results expected to be reached by

digital pointing devices primarily used for presenta-

tions. For a more precise assessment of the device

properties, the determined factors will probably have

to be adapted to result in scores that are more true to

the subjective user assessments, which increases the

reliability of the proposed testing methods.

The adaption of the empirically determined fac-

tors leads us to the outlook for further development.

First of all the subjective user assessment has to be

enhanced so that the device tester not only rank the

tested devices from best to worst but also rate them

on a scale from 0 to 10 for each of the three proper-

ties as well as for the overall usability. Based on the

results from the subjective assessment the calculation

of the score from the tests should be adapted to be as

similar to the user rating as possible.

REFERENCES

Atkinson, P. (2006). The best laid schemes o’ mice and

men : the evolution of the computer mouse. In De-

sign and Evolution : Proceedings of Design History

Society Conference.

Batra, S., Dykstra, D., Hsu, P., Radle, K. A., and Wieden-

beck, S. (1998). Pointing device performance for lap-

top computers. In Proceedings of the Human Factors

and Ergonomics Society Annual Meeting, volume 42-

6, pages 536–540. SAGE Publications Sage CA: Los

Angeles, CA.

Charness, N., Holley, P., Feddon, J., and Jastrzembski, T.

(2004). Light pen use and practice minimize age and

hand performance differences in pointing tasks. Hu-

man Factors, 46(3):373–384.

Douglas, S. A., Kirkpatrick, A. E., and MacKenzie, I. S.

(1999). Testing pointing device performance and user

assessment with the iso 9241, part 9 standard. In Pro-

ceedings of the SIGCHI conference on Human Factors

in Computing Systems, pages 215–222. ACM.

Greenemeier, L. (2009). The origin of the com-

puter mouse. Website. Online available at

https://www.scientiﬁcamerican.com/article/origins-

computer-mouse accessed at 14-06-2019.

ISO (2007). 9241-400: Ergonomics of human–system in-

teraction – part 400: Principles and requirements for

physical input devices.

MacKenzie, I. S. and Jusoh, S. (2001). An evaluation of

two input devices for remote pointing. In IFIP In-

ternational Conference on Engineering for Human-

Computer Interaction, pages 235–250. Springer.

MacKenzie, I. S., Kauppinen, T., and Silfverberg, M.

(2001). Accuracy measures for evaluating computer

pointing devices. In Proceedings of the SIGCHI

conference on Human factors in computing systems,

pages 9–16. ACM.

Oehl, M., Sutter, C., and Zieﬂe, M. (2007). Considerations

on efﬁcient touch interfaces–how display size inﬂu-

ences the performance in an applied pointing task. In

Symposium on Human Interface and the Management

of Information, pages 136–143. Springer.

Multi-aspect Evaluation Method for Digital Pointing Devices

135