Conference PaperPDF Available

Using Eye Tracking and Task Analysis for Analysing Usability Interactions

Authors:

Abstract and Figures

The objective of this study is to characterize and analyse how affordances of packages are perceived by users and what type of behaviour affordances trigger. In order to characterize the effect of individual design features on user attention and behaviour, a commercially available package for epinephrine was redesigned and tested with a panel of participants (n=39). Each experiment was conducted as a randomized complete block design with three factors: location of opening mechanism (top or bottom), type of opening mechanism (press-in or pull tab), and colour contrast in opening area (with or without). By crossing all possible conditions (2x2x2), eight different folding carton designs resulted. Each participant was treated as a block and presented all eight designs in a random order. Participants stood behind a counter of a fixed height wearing an ASL Mobile 5 glasses and completed eight opening tasks in a lab facility. They were instructed to imagine an emergency scenario where they needed to remove all contents from each package as quickly as possible. Each opening trial was analysed using task analysis (TA) and eye tracking analysis (EA). TA re-lied on video feeding from the eye-tracker's scene camera and eye movement information to divide the human-package interaction in five subtasks (i.e., reaching package, scanning for opening mechanism, improper attempt to open, proper attempt to open, and removing contents). EA used eye-movement data and three areas of interest (i.e., package's opening end, package's body text, package's non-opening end) to determine; overall mean fixation duration, overall number of fixations, number of fixations per AOI, total dwell duration per AOI, percentage of total dwell duration per AOI. The TA revealed significant differences (p<0.008) in time spent in four subtasks (i.e., scanning for opening mechanism, improper attempt to open, proper attempt to open, and reaching for con-tents) when comparing designs with a pull tab and designs with press-in mechanisms. TA also revealed significant differences (p<0.008) in time spent in two subtasks (i.e., scanning for opening mechanism, improper attempt to open) when comparing designs with an opening mechanism at the top and designs with an opening mechanism at the bottom of the package. The TA revealed no significant differences (p<0.008) for colour contrast (with and without) in the opening area. The EA shed light about the efficiency of different design factors in grabbing the participants' at-tention and conveying information. Significant differences (p<0.008) were found in overall number of fixations when comparing designs with a pull tab and designs with a press-in mechanism. A closer look at the number of fixations per AOI, revealed significant differences (p<0.008) in num-ber of fixations at package's non-opening end when comparing designs with a pull tab and de-signs with press-in mechanism. EA also revealed significant differences (p<0.008) in total dwell duration at packages' non-opening end when comparing two design factors; opening mechanism (pull tab vs press-in mechanism), and opening location (top vs bottom). Further analysis comparing percentage of total dwell time spent on each AOI revealed significant differences in packages' opening end and non-opening end when comparing two design factors; opening mechanism, and opening location. All these findings have critical implications to understand how packaging design can guide cognitive efforts effectively towards areas that accurately deliver information to fulfill a task.
Content may be subject to copyright.
28th IAPRI World Symposium on Packaging
503
Using Eye Tracking and Task Analysis for Analysing
Usability Interactions!
Irene Carbonell1, Javier de la Fuente1*, Trevor Lowe1, Jordan Hill1, Soma Roy2, Colleen
Twomey3
1 Industrial Technology and Packaging, California Polytechnic State University, United States
2 Statistics Department, California Polytechnic State University, United States
3 Graphic Communication, California Polytechnic State University, United States
*Corresponding author name. Email: jdelafue@calpoly.edu
Keywords: package design, eye tracking, usability, affordances, medical device packaging, ergonom-
ics, task analysis.
Abstract: The objective of this study is to characterize and analyse how affordances of packages are
perceived by users and what type of behaviour affordances trigger. In order to characterize the effect
of individual design features on user attention and behaviour, a commercially available package for
epinephrine was redesigned and tested with a panel of participants (n=39). Each experiment was
conducted as a randomized complete block design with three factors: location of opening mechanism
(top or bottom), type of opening mechanism (press-in or pull tab), and colour contrast in opening area
(with or without). By crossing all possible conditions (2x2x2), eight different folding carton designs re-
sulted. Each participant was treated as a block and presented all eight designs in a random order.
Participants stood behind a counter of a fixed height wearing an ASL Mobile 5 glasses and completed
eight opening tasks in a lab facility. They were instructed to imagine an emergency scenario where
they needed to remove all contents from each package as quickly as possible.
Each opening trial was analysed using task analysis (TA) and eye tracking analysis (EA). TA relied on
video feeding from the eye-tracker’s scene camera and eye movement information to divide the hu-
man-package interaction in five subtasks (i.e., reaching package, scanning for opening mechanism,
improper attempt to open, proper attempt to open, and removing contents). EA used eye-movement
data and three areas of interest (i.e., package’s opening end, package’s body text, package’s non-
opening end) to determine; overall mean fixation duration, overall number of fixations, number of fixa-
tions per AOI, total dwell duration per AOI, percentage of total dwell duration per AOI.
The TA revealed significant differences (p<0.008) in time spent in four subtasks (i.e., scanning for
opening mechanism, improper attempt to open, proper attempt to open, and reaching for contents)
when comparing designs with a pull tab and designs with press-in mechanisms. TA also revealed sig-
nificant differences (p<0.008) in time spent in two subtasks (i.e., scanning for opening mechanism,
improper attempt to open) when comparing designs with an opening mechanism at the top and de-
signs with an opening mechanism at the bottom of the package. The TA revealed no significant differ-
ences (p<0.008) for colour contrast (with and without) in the opening area.
The EA shed light about the efficiency of different design factors in grabbing the participants’ attention
and conveying information. Significant differences (p<0.008) were found in overall number of fixations
when comparing designs with a pull tab and designs with a press-in mechanism. A closer look at the
number of fixations per AOI, revealed significant differences (p<0.008) in number of fixations at pack-
age’s non-opening end when comparing designs with a pull tab and designs with press-in mechanism.
EA also revealed significant differences (p<0.008) in total dwell duration at packages’ non-opening
end when comparing two design factors; opening mechanism (pull tab vs press-in mechanism), and
opening location (top vs bottom). Further analysis comparing percentage of total dwell time spent on
each AOI revealed significant differences in packages’ opening end and non-opening end when com-
paring two design factors; opening mechanism, and opening location.
All these findings have critical implications to understand how packaging design can guide cognitive
efforts effectively towards areas that accurately deliver information to fulfil a task.
504
1 Introduction
Numerous research has been conducted on the use of eye tracking technology to investigate cognitive
processes. This is because eye movements reflect the human thought processes; so the observer’s
thought may be followed from records of eye-movement [1]. Therefore, by measuring eye movements,
one can find functional links between what is observed and cognitive processing of the particular ob-
ject. Studies have focused on multiple and diverse areas of academic research.
Extensive eye movement data has proven useful, for example, in studying reading tasks [2][5], scene
perception [6], and computer interface evaluation [7]. Use of eye-tracking technology for human fac-
tors research has also been conducted in a wide-ranging set of areas, including automotive [8], [9]
aviation [10], [11], and household products [12], [13]. Eye tracking research on human perception of
product design using a screen image of the product has also been reported [14], [15].
In packaging research, eye tracking technology has been used for measuring the effect of package
design on shelf impact [16], warning noticeability [17], and processing of written information [18].
However, there are no reports on using this technology to understand how structural and graphic de-
sign affect package usability.
This study proposes the use of task analysis and eye tracking analysis to investigate the effect that
various design affordances have on the ability for users to locate and understand opening mecha-
nisms of a package. The package selected to develop this methodology is a folding carton package
found in emergency room crash carts. A crash cart is a wheeled cabinet with drawers used in hospi-
tals for transporting and dispensing emergency medication and equipment during life support proto-
cols (Figures 1 and 2). These crash cart drugs are used in contexts where high emotional stress and
fast pace are the norm. Most of these drugs are packaged in folding cartons with opening mechanisms
that involve press-in and tearing on the bottom of the package (Figure 3). Previous research by de la
Fuente et al. [19], [20] suggests that these packages are counterintuitive for lay users and healthcare
professionals. They also showed a relationship between design affordances, opening times, and
opening errors. The present study aimed to explain the reasons behind this relation and to provide
data that supports the importance of affordances in usability.
!
Figure 1: Crash
cart.
Figure 2: A crash cart drugs’ tray at Sierra
Vista Regional Medical Center,
San Luis Obispo, CA.
Figure 3: A typical press-in
opening mechanism for crash
cart drugs packaged in folding
cartons.
505
2 Materials and Methods
In order to measure the effect of individual folding carton design features on several eye-tracking
measures and time spent on opening subtasks, a commercially available crash cart drug package was
redesigned and tested with lay users. The experiment was conducted as a randomized complete block
design with three factors: location of opening mechanism (top or bottom of the package), type of open-
ing mechanism (pull tab or press-in and tear), and colour contrast in opening area (with and without
contrast). The existing standard package utilizes a press-in and tear on the bottom of the package
without a colour contrast. Crossing the three design factors resulted in a 2x2x2 design of experiment,
resulting in eight different folding carton designs (Table 1). Each participant was treated as a block,
and all eight designs were presented to each participant in a random order to ensure an equal distribu-
tion of first exposure of each package design.
Table 1: Characteristics of the eight package design variations tested.
Opening
Mechanism
Opening
Location
Coloured
Opening
Press-in
Bottom
No
Press-in
Top
No
Press-in
Bottom
Yes
Press-in
Top
Yes
Pull tab
Bottom
No
Pull tab
Top
No
Pull tab
Bottom
Yes
Pull tab
Top
Yes
2.1 Participants
Using procedures approved by Cal Poly’s Human Subjects Committee, participants were recruited on
the campus of California Polytechnic State University (San Luis Obispo, CA). Prior to testing, re-
searchers gathered information from each participant regarding gender, age, preferred hand, visual
acuity, colour blindness, and whether or not they had professional experience opening packages con-
taining crash cart drugs. Each participant’s visual acuity was determined using a Snellen chart from
Eye Chart (Dok LLC). Subjects stood 4 feet away from the chart posted on a wall and were asked to
read the lowest line possible. Ishihara plates number one and two were used to evaluate participants’
colour deficiencies [21].
2.2 Packages
Three hundred and twelve folding cartons (39 participants × 8 designs) were printed on white paper-
board (coated SBS 18 points, 100 pounds) in a production printer (Konica Minolta C1100 BizHub, To-
kyo, Japan), and cut/creased on a computer-controlled cutting and creasing table (Kongsberg i-XE1
by Esko-Graphics bvba, Ghent, Belgium). The 8 designs were designed using illustration software (Il-
lustrator® CC by Adobe®, San Jose, CA, USA). Names for a fictitious pharmaceutical company (i.e.,
Pharm®) and a non-existent drug (i.e., Tyvedron®) were created to make the designs more credible
and realistic to the user. The combination of three design factors resulted in eight different designs
(Figure 4). The colour contrast combination (i.e., light blue and dark orange) selected can be per-
ceived by subjects with total colour blindness.
506
Figure 4: Eight package design variations.
2.3 Experimental Setup
2.3.1 Eye-tracking Equipment
Participants’ eye movements and scene video from their point of view were recorded using a mobile
eye-tracking device (MobileEye-5 by Applied Science Laboratories Inc., Bedford, MA, USA). The cali-
bration sequence employed a letter-size calibration board with five dots distributed across its surface,
and placed on the testing counter. This board was used before each trial to verify calibration accuracy.
2.3.2 Setup and Tasks
Participants stood behind a table (90 cm) and completed eight opening tasks in a lab facility. Partici-
pants were instructed to imagine an emergency scenario where they needed to remove all contents
from each package as quickly as possible. Plastic trays were placed over each package so partici-
pants could not see them until instructed (Figure 5). Packages were positioned so that the longest di-
mension of the folding carton was horizontal and the drug name was on the left. Participants were in-
structed to remove all contents (one PVC tube and one wooden dowel whose weight and size emulat-
ed an epinephrine-type product) from the package as quickly as possible and to place the contents in
a plastic tray. In between opening tasks, participants spent one minute playing a game on a tablet
computer.
Figure 5: Testing area setup.
507
2.4 Data Analysis
Statistical analyses were performed using JMP® [22]. Analyses of variance (ANOVA) were conducted
to investigate whether and how variables related to package design features. Each opening trial was
analysed using ASL Results Plus [23] to perform a task analysis and an eye tracking analysis.
2.4.1 Task Analysis (TA)
TA relied on video from the eye-tracker’s scene camera to divide the human-package interaction in
subtasks. Ten events were coded to perform time calculations (Table 2). Using these events, five sub-
tasks were calculated:
Reaching: Time from when package is first visible in its entirety until time package is first touched.
Scanning for opening mechanism: Time from when package is first touched until time subject at-
tempts to open package.
Improper attempt to open: Time from when subject attempts to open package at non-opening end
(NE) until time subject opens package at NE, or abandons attempt.
Proper attempt to open: Time from when subject attempts to open package at opening end (OE)
until time subject opens package at OE, or abandons attempt.
Removing contents: Time from when subject opens package until time subject places contents in
basket.
Calculations of the time spent on each subtask are conditional and depend on the sequence of events
(Table 3). Total opening time was defined as the time participants took to understand the package,
find the opening mechanism, open the package, and remove its contents, or in other words, time sub-
ject spent on all five subtasks.
Table 2: Coded events.
EVENTS
CODE
Visible
V
Touch
T
Attempt opening at NE
ANE
Package opened at NE
ONE
Abandon attempt at NE
ABNE
First dowel in basket
D1
Attempt opening at OE
AOE
Package opened at OE
OOE
Abandon attempt at OE
ABOE
All dowels in basket
D2
508
Table 3: Subtasks’ time calculations.
SUBTASKS
CALCULATION
DESCRIPTION
Reaching
T - V
Time from when package is first visible in its entirety until time
package is first touched
Scanning
ANE - T
Time from when package is first touched until time subject
begins to open package at NE
AOE - T
Time from when package is first touched until time subject
begins to open package at OE
(ANE - T) + (AOE - ONE)
(Time from when package is first touched until time subject
begins to open package at NE) + (Time from when package is
opened at NE until time subject attempts to open package at
OE)
(ANE - T) + (AOE - ABNE)
(Time from when package is first touched until time subject
begins to open package at NE) + (Time from when subject
abandons attempt to open package at NE until time subject
attempts to open package at OE)
(ANE - T) + (AOE - D1)
(Time from when package is first touched until time subject
begins to open package at NE) + (Time from when first dowel
is in basket until time subject attempts to open package at
OE)
Improper
attempt
0
There is no improper attempt to open package
ONE - ANE
Time from when subjects attempts to open package at NE
until time subject opens package at NE
ABNE - ANE
Time from when subjects attempts to open package at NE
until time subject abandons attempt at NE
Proper
attempt
0
There is no proper attempt to open package
OOE - AOE
Time from when subject attempts to open package at OE until
time subject opens package at OE
ABOE - AOE
Time from when subject attempts to open package at OE until
time subject abandons attempt at OE
Removing
contents
D2 - ONE
Time from when subject opens package at NE until time sub-
ject places second dowel in basket
D2 - OOE
Time from when subject opens package at OE until time sub-
ject places second dowel in basket
(D1 - ONE ) + (D2 - OOE)
(Time from when subject opens package at NE until time sub-
ject places first dowel in basket) + (Time from when subject
opens package at OE until time subject places second dowel
in basket)
509
2.4.2 Eye Tracking Analysis (EA)
Eye tracking videos were analysed to determine fixations and dwell in relation to predefined areas of
interest. Fixation is the state in which the eye remains static over a period of time. A maximum visual
angle variation and a minimum time have to be defined to consider a fixation has started. For this
study, they were defined as 1 degree and 100 milliseconds. Areas of interest (AOIs) are regions the
researcher is interested in gathering data about. For this study, we defined three areas of interest:
Opening End (OE), Main Body (MB), and Non-opening End (NE) of package (Figure 6). Dwell
measures how long gaze remains inside an AOI. The properties of five commonly used measures
were used to extract useful information:
Overall average fixation duration
Overall number of fixations
Number of fixations per AOI
Total dwell duration on each AOI
Percentage of total dwell time on each AOI
Figure 6: Screen capture of ASL Results Plus software showing current point of gaze (green cross-
hair), gaze trail (red circles), fixations (green circles), and the three areas of interest (AOIs): Opening
End (dark red), Main Body (orange), and Non-opening End (blue).
3 Results
3.1 Participants
A total of 39 participants tested the eight folding carton designs. None of the participants had profes-
sional experience using folding carton packages for vial and syringe medical products found in crash
carts. Four participants were discarded due to technical issues with equipment and/or packages. Par-
ticipants’ visual acuity was distributed as follows: 20/15 (44%), 20/20 (38%), 20/25 (8%), 20/30 (5%),
20/40 (3%), and 20/50 (3%). None of the participant had total colour blindness; two participants (5%)
exhibited signs of red-green deficiencies.
3.2 Task Analysis (TA)
A total of 23 participants were used to perform analyses of variance. These ANOVAs accounted for
2x2x2 factorial treatment structure and randomized complete block design. A bar chart for time spent
on the five subtasks for each design variation is shown in Figure 7 and Figure 8; all subtasks’ times
were adjusted and normalized to account for person-to-person variability.
510
Separate ANOVA models were run for each of the following response variables, and the same ex-
planatory variables: opening location, opening mechanism, coloured opening, and two-way interaction,
as well as subject identification as the random effect to account for user variability. TA results for each
subtask revealed the following:
a) Reaching: There was no evidence for an effect of any of the design factors in reaching time.
b) Scanning for opening mechanism: Having a pull tab opening mechanism significantly reduced
scanning time (p<0.0001). Having the opening mechanism on the top of the package also signifi-
cantly reduced scanning time (p<0.0001). More specifically, using Tukey’s HSD method, compar-
ing the location of the press-in mechanism (bottom or top), shows packages having a press-in
mechanism at the bottom rather than the top exhibit significantly increased scanning times
(p<0.0001).
c) Improper attempt to open: Having a pull tab opening mechanism significantly reduced time spent
in improper attempt to open (p<0.0001). Having the opening mechanism on the top of the package
significantly reduced the time spent in improper attempt to open (p=0.0002). More specifically, us-
ing Tukey’s HSD method, comparing the location of the press-in mechanism (bottom or top),
shows packages having a press-in mechanism at the bottom rather than the top exhibit significant-
ly increased improper attempt to open times (p<0.0075).
d) Proper attempt to open: Packages with a pull tab showed statistically significantly shorter times for
trying to open the package from the opening end (p<0.0001).
e) Removing contents: Packages with pull tab mechanism showed statistically significantly shorter
times for removing contents from the package (p=0.0047). There was no evidence for an effect of
opening location or colour contrast.
Overall, the use of a pull tab and the top opening location significantly increased users’ ability to locate
the opening mechanism as well as reduced time wasted attempting to open the package at an im-
proper location. Additionally, the pull tab facilitated a faster total opening time, granting users access
to the contents sooner.
Figure 7: Segmented bar chart of Task Analysis for each package design.
Opening times have been adjusted for person-to-person variability.
Means not sharing a letter are significantly different (p<0.008).
511
Figure 8: Segmented bar chart of Task Analysis for each package design.
Opening times have been adjusted for person-to-person variability.
Means not sharing a letter are significantly different (p<0.008).
3.3 Eye Tracking Analysis (EA)
Analyses of variance were run on JMP, focusing on the first package trial of 24 subjects to control
the effect that previous experiences might have on the outcome.
3.3.1 Overall average fixation duration
We could not find evidence for an effect of opening mechanism, opening location, or colour on overall
average fixation duration. Longer fixation duration is often associated with a deeper, more effortful
cognitive processing. In usability research, this generally indicates difficulty in extracting information
from the object.
3.3.2 Overall average number of fixations
Packages with a pull tab showed statistically significantly fewer fixations (p=0.0007). There was no
evidence of an effect of opening location or colour. The overall number of fixations is thought to be
negatively correlated with the efficiency of given tasks in usability research [6]. Therefore, having a
pull tab increased the efficiency of opening the package by reducing the overall number of fixations
required. A bar chart for overall average number of fixations per design factor is shown on Figure 9.
512
3.3.3 Number of fixations per AOI
Opening End (OE): There was no evidence of an effect of opening mechanism, opening location, or
colour, in the number of fixations on OE.
Main Body (MB): There was no evidence of an effect of opening mechanism, opening location, or
colour, in the number of fixations on MB.
Non-Opening End (NE): There was no evidence for an effect of colour contrast on number of fixa-
tions on NE, but having a press-in opening mechanism significantly increased the number of fixa-
tions on the NE (p=0.0026). More specifically, using Tukey’s HSD method, comparing the location of
the pull tab (bottom or top) shows no significance difference in regards to number of fixations on NE.
However, having a press-in mechanism at the top rather than the bottom shows significant reduction
in the number of fixations on NE (p=0.0033).
The number of fixations on a given area could reflect its importance [7]. These results show that, for
designs that had a press-in opening mechanism located at the bottom of the package, subjects con-
sidered the non-opening end of the package to be important to fulfil their task.
3.3.4 Total dwell duration per AOI
Opening End (OE): There was no evidence of an effect of opening mechanism, opening location, or
colour contrast in the total dwell duration on the OE.
Main Body (MB): There was no evidence of an effect of opening mechanism, opening location, or
colour contrast in the total dwell duration on the MB.
Non-Opening End (NE): Packages with a press-in opening mechanism exhibited statistically signifi-
cantly longer dwell times on the NE (p=0.0007) than packages with a pull tab mechanism. Having a
bottom opening location as compared to a top opening location also resulted in significantly in-
creased total dwell duration on the NE (p=0.0021). More specifically, using Tukey’s HSD method,
comparing the location of the press-in mechanism (bottom or top) shows having a press-in mecha-
nism at the bottom rather than the top shows significantly increased total dwell duration on the NE
(p=0.0003). There was no evidence for an effect of colour contrast on the total dwell duration on NE.
Longer total dwell duration on a given AOI could indicate difficulty in extracting or interpreting infor-
mation from that area [7]. These results show that, for designs that had a press-in opening mechanism
and/or the opening mechanism located at the bottom of the package, subjects spent a significant
amount of time trying to understand how to open the package from the non-opening end.
3.3.5 Percentage of total dwell time per AOI
Opening End (OE): Having a press-in opening mechanism significantly decreased the percentage
total dwell time on the OE (p=0.0018). Having the opening location at the bottom of the package
significantly decreased the total dwell time on the OE (p=0.0051). More specifically, using Tukey’s
HSD method, comparing the location of the press-in mechanism (bottom or top) shows having a
press-in mechanism at the bottom rather than the top shows significantly decreased percentage of
total dwell time spent on OE (p=0.0002). There was no evidence for an effect of colour contrast on
the percentage of total dwell time on OE.
Main Body (MB): There was no evidence for an effect of opening mechanism, opening location, or
colour contrast in the percentage of total dwell time spent on the MB.
Non-Opening End (NE): Having a press-in opening mechanism significantly increased the percent-
age of total dwell time spent on the NE (p<0.0001). Having the opening location at the bottom of the
package significantly increased the percentage of total dwell time spent on the NE (p<0.0001). More
specifically, using Tukey’s HSD method, comparing the location of the press-in mechanism (bottom
or top) shows having a press-in mechanism at the bottom rather than the top shows significantly in-
creased percentage of total dwell time spent on NE (p<0.0001). There was no evidence for an effect
of colour contrast on the percentage of total dwell time on NE.
The proportion of time looking at a particular AOI could reflect the importance the subjects gives to this
particular AOI. A line graph for percentage of total dwell time per AOI is shown on Figure 10. These
results show that packages with a press-in opening mechanism and/or a bottom opening location will
increase the perceived importance of the non-opening end, while reducing the perceived importance
of the opening end, to fulfil the task. Figure 11 shows percentage of total dwell time per AOI and de-
513
sign factor.
Figure 9:
Overall average number of fixations per design factor
* indicates significant difference (p<0.008)
Figure 10
: Percentage of total dwell time per AOI.
514
Figure 11
: Percentage of total dwell time per AOI and design factor.
* indicates significant difference (p<0.008)
515
4 Conclusions
This study explored the use of two complementary analyses to understand human-packaging interac-
tions. The use of task analysis and eye tracking analysis proved to be valuable in identifying how
packaging design features can significantly improve the efficiency of common tasks involved in the
use of a package. The following remarks are possible interpretations of the analysed measures and
their meaning in the context of usability testing:
Task Analysis
Packages with affordances that clearly communicate opening regions exhibit shorter scanning task
times as well as reduced time spent attempting to open at an improper location.
Package designs that are easy to understand and operate will reduce time spent attempting to open
package at proper location.
Eye Tracking Analysis
Packages with strong affordances (e.g., having a very visible tab) will likely result in interactions in
which the overall number of fixations is minimized, thus creating a more efficient (e.g., faster, with
less errors, etc.) and pleasant experience.
Number of fixations per AOI can be used as an indicator of how well the package is communicating
its affordances. Packages that do not communicate well and contain misleading affordances can di-
rect user’s attention away from the area of interest needed to fulfil a task.
Total dwell duration on AOIs provide evidence about how a poor package design can guide sub-
ject’s cognitive efforts towards areas with non-relevant information for the task at hand.
Percentage of total dwell duration on AOIs can shed light about the role of design features and their
perceived importance. An increase of the perceived importance of non-relevant areas will reduce the
perceived importance of relevant areas.
5 Acknowledgements
The authors would like to thank the participants who took part in the experiments, the National Center
for Patient Safety who initially suggested this issue as a healthcare problem, and the Orfalea College
of Business for funding this project through the Interdisciplinary Research Grant Program.
6 References
[1] A. L. Yarbus, Eye Movements and Vision. New York: Plenum Press, 1967.
[2] M. Dambacher and R. Kliegl, “Synchronizing Timelines: Relations Between Fixation Durations
and N400 Amplitudes During Sentence Reading,” Brain Res., vol. 1155, no. 1, pp. 147–162,
2007.
[3] K. Rayner, “Eye Movements in Reading and Information Processing: 20 Years of Research,”
Psychol. Bull., vol. 124, no. 3, pp. 372–422, 1998.
[4] K. Rayner and A. Pollatsek, The Psychology of Reading. Englewood Cliffs, NJ: Prentice Hall,
1989.
[5] M. Just and P. Carpenter, “A Theory of Reading: From Eye Fixations to Comprehension,”
Psychol. Rev., vol. 87, no. 4, pp. 329–354, 1980.
[6] J. M. Henderson, P. A. Weeks, and A. Hollingworth, “The Effects of Semantic Consistency on
Eye Movements During Complex Scene Viewing,” J. Exp. Psychol. Hum. Percept. Perform., vol.
25, no. 1, pp. 210228, 1999.
[7] J. H. Goldberg and X. P. Kotval, “Computer Interface Evaluation Using Eye Movements: Methods
and Constructs,” Int. J. Ind. Ergon., vol. 24, no. 6, pp. 631645, 1999.
[8] O. Laya, “Eye Movements in Actual and Simulated Curve Negotiation Tasks,” Int. Assoc. Traffic
Saf. Sci. Res., vol. 16, no. 1, pp. 15–26, 1992.
[9] D. Shinar, E. D. McDowell, and T. H. Rockwell, “Eye Movements in Curve Negotiation.,” Hum.
Factors, vol. 19, no. 1, pp. 63–71, 1977.
516
[10] R. L. Harris and D. M. Christhilf, “What Do Pilots See in Displays?,” in 24th Annual Meeting of the
Human Factors Society, 1980, pp. 2226.
[11] M. J. Fitts PM, Jones RE, “Eye Movements of Aircraft Pilots During Instrument-Landing
Approaches,” Aeronaut. Eng. Rev., vol. 9, no. 2, pp. 24–29, 1950.
[12] L.-H. Chen, C.-F. Lee, and C.-M. Tsai, “Perceiving Affordances Through Perceptual Information,”
in International Association of Societies of Design Research, 2007, pp. 18.
[13] M. Land, N. Mennie, and J. Rusted, “The Roles of Vision and Eye Movements in the Control of
Activities of Daily Living,” Perception, vol. 28, no. 11, pp. 1311–1328, 1999.
[14] K. Koivunen, S. Kukkonen, and S. Lahtinen, “Towards Deeper Understanding of How People
Perceive Design in Products,” in CADE2004 Web Proceedings of Computers in Art and Design
Education Conference, 2004, pp. 25.
[15] T. Reid, E. MacDonald, and P. Du, “Impact of Product Design Representation on Customer
Judgment,” J. Mech. Des., vol. 135, no. 091008, pp. 1–12, 2012.
[16] J. Clement, “Visual Influence on In-Store Buying Decisions: an Eye-track Experiment on the
Visual Influence of Packaging Design,” J. Mark. Manag., vol. 23, no. 9–10, pp. 917928, 2007.
[17] A. Gawasane, L. Bix, J. de la Fuente, R. P. Sundar, and T. J. Smith, “Consumer Attention to an
Over-the-counter Warning in Four Different Styles of Design,” Packag. Technol. Sci., vol. 25, no.
7, pp. 385–396, Nov. 2012.
[18] J. H. Goldberg, C. K. Probart, and R. E. Zak, “Visual Search of Food Nutrition Labels,” Hum.
Factors, vol. 41, no. 3, pp. 425–437, 1999.
[19] J. de la Fuente, S. Gustafson, C. Twomey, and L. Bix, “An Affordance-Based Methodology for
Package Design,” Packag. Technol. Sci., vol. 28, no. 2, pp. 157–171, Feb. 2015.
[20] J. de la Fuente et al., “Measuring the Effect of Affordances on a Crash Cart Medicine Packaging,”
in 27th IAPRI Symposium - International Association of Packaging Research Institutes, 2015, pp.
394403.
[21] S. Ishihara, The Series of Plates Designed as a Test for Colour-Blindness, 24 Plates. Tokyo,
Japan: Kanehara Shuppan Co., Ltd., 1972.
[22] SAS Institute Inc., “JMP® Pro v14.0.0 for Mac.” SAS Institute Inc., Cary, North Carolina, USA,
2016.
[23] Applied Science Group LLC, “ASL Results Plus v1.8.3.15 for Windows.” Bedford, Massachusets,
USA.
... This dynamic pattern is attributed to two modes of acquiring information: exploration and inspection. In packaging research, eye tracking technology has been used for measuring the effect of package design on shelf impact [9] , warning noticeability [10] , processing of written information [11] , opening time and opening errors [12] . However, there are no reports on using eye movement behaviour to understand how structural and graphic design affect package usability. ...
... In order to measure the effect of task difficulty on eye movement behaviour, a commercially available crash cart drug package was redesigned and tested with lay users. Previous research by de la Fuente et al. [14], [15] and Carbonell et al. [12] showed a relationship between design affordances and task difficulty. The present study uses the folding carton designs that scored highest and lowest in task difficulty. ...
Conference Paper
Full-text available
Mobile eye-tracking technology provides data that can be linked to user attention during visual perception. The relationship between a fixation and its subsequent saccade has been proposed to identify ambient versus focal visual processing. Short fixation durations followed by long saccades are characteristic of ambient processing. Longer fixation durations followed by shorter saccades suggests focal processing. The objective of this study was to apply a novel parametric scale (i.e., coefficient K) to eye movement data from packaging usability trials of known difficulty to investigate the relationship between fixations, saccades, and cognitive-challenging phases in user-package interactions. A panel of 40 participants (20 females, 20 males) were asked to open a folding carton and extract its contents. Each participant stood behind a counter of a fixed height wearing an ASL Mobile 5 glasses and completed one task in a lab facility. They were instructed to imagine an emergency scenario where all package contents had to be removed as quickly as possible. Half of the participants were randomly assigned a package with a pull tab on the top of the container (i.e., easy task) while the other half tested a package with a press-in mechanism on the bottom of the folding carton (i.e., difficulty task). Coefficients for each trial were calculated using fixation durations and either saccade durations or saccade amplitudes. Two-sample t-tests based on the coefficients K were conducted to compare mean differences across the two tasks. For coefficients K calculated using saccade durations, results revealed a significant difference in mean coefficient K for each type of task (t(27)=2.57, p≤0.05). For coefficients K calculated using saccade amplitudes, results revealed a significant difference in mean coefficient K for each type of task (t(34)=2.72, p≤ 0.01). Moreover, mean coefficients K for the easy task were negative, signalling ambient processing, while mean coefficients K for the difficult task were positive, an indication of focal processing. Gender differences are also reported. Findings have critical implications for understanding user-package interaction problems and support the use of algorithmic processing of eye-tracking data for automatic usability problems detection.
Conference Paper
Full-text available
In the United States, most drugs stored in crash carts of emergency rooms are packaged in folding cartons with opening mechanisms that involve pressing and tearing the bottom of the package. Anecdotal evidence and a previous study conducted by the research team suggest that these packages are counterintuitive for lay users. The concept of affordances, how design features communicate actionable possibilities, can be applied to improve the usability of packaging. In order to measure the effect of individual design features and previous experience on opening time and error frequency during first opening attempt, a commercially available package for epinephrine was redesigned and tested with two panels of participants; 33 lay users and 17 healthcare providers. Each experiment was conducted as a randomized complete block design with three factors: location of opening mechanism (top or bottom), type of opening mechanism (press-in or tab), and colour contrast in opening area (with and without). By crossing all possible conditions (2x2x2), eight different folding carton designs resulted. Each participant was treated as a block and presented all eight designs in a random order. Participants stood behind a counter of a fixed height and completed eight opening tasks in a lab facility (lay users) and an emergency room (healthcare providers). They were instructed to imagine an emergency scenario where they needed to remove all contents from each package as quickly as possible. Results show that colour contrast had no significant effect on opening time, having a tab significantly reduces time to open. More specifically, tabs cued the user as to the correct end containing the opening feature regardless of it was positioned on the top or bottom of the package. When there was no tab, then having the opening at the top resulted in a significant reduction in time to open, compared to having the opening at the bottom. In an analysis comparing the eight designs, the actual commercial packages ranked as the worst designs in terms of opening time and error frequency. Findings have critical implications for designing packages that are more usable and for eliminating errors during product use.
Article
Full-text available
When researchers ask customers to judge product form during the design process, they often manipulate simplified product representations, such as silhouettes and sketches, to gather information on which designs customers prefer. Using simplified forms, as opposed to detailed realistic models, make the analysis of gathered information tractable and also allows the researcher to guide customer focus. The theory of constructed preferences from psychology suggests that the product form presented will influence customer judgments. This paper presents a study in which subjects were shown computer sketches, front/side view silhouettes, simplified renderings, and realistic renderings to test the extent to which a variety of judgments including opinions, objective evaluations, and inferences are affected by form presentation. Results show a variety of phenomena including preference inconsistencies and ordering effects that differed across type of judgment. For example, while inferences were consistent across form, opinions were not. An eye tracker identified differences in viewing strategies while making decisions. Associated data, such as fixation times and fixation counts, provide additional insight into findings.
Article
Full-text available
The term affordance describes an object's utilitarian function or actionable possibilities. Product designers have taken great interest in the concept of affordances because of the bridge they provide relating to design, the interpretation of design and, ultimately, functionality in the hands of consumers. These concepts have been widely studied and applied in the field of psychology but have had limited formal application to packaging design and evaluation. We believe that the concepts related to affordances will reveal novel opportunities for packaging innovation. To catalyse this, presented work had the following objectives: (a) to propose a method by which packaging designers can purposefully consider affordances during the design process; (b) to explain this method in the context of a packaging-related case study; and (c) to measure the effect on package usability when an affordance-based design approach is employed.
Article
Full-text available
This article describes the impact of visual attention on consumers' in-store buying behaviour. Through an eye-track experiment, it demonstrates the advantage of a behaviour model that addresses visual attention and an increase in visual stimuli during the process. It reveals that consumers exhibit a muddled search strategy where packaging design influences the decision process in several phases. Five phases were found in an in-store decision process, and the post-purchase phase seems to be essential for even low-level in-store decision processes. Further knowledge on packaging design elements is needed for a broader understanding of visual influence during in-store purchase decisions.
Article
Full-text available
Eye movements were recorded while participants viewed line-drawing pictures of natural scenes in preparation for a memory test (Experiment 1) or to find a target object (Experiment 2). Initial saccades in a scene were not controlled by semantic information in the visual periphery, although fixation densities and fixation durations were affected by semantic consistency. The results are compared with earlier eye-tracking studies, and a qualitative model of eye movement control in scene perception is discussed in which initial saccades in a scene are controlled by visual but not semantic analysis. (PsycINFO Database Record (c) 2012 APA, all rights reserved)
Article
Full-text available
Eye movement-based analysis can enhance traditional performance, protocol, and walk-through evaluations of computer interfaces. Despite a substantial history of eye movement data collection in tasks, there is still a great need for an organized definition and evaluation of appropriate measures. Several measures based upon eye movement locations and scanpaths were evaluated here, to assess their validity for assessment of interface quality. Good and poor interfaces for a drawing tool selection program were developed by manipulating the grouping of tool icons. These were subsequently evaluated by a collection of 50 interface designers and typical users. Twelve subjects used the interfaces while their eye movements were collected. Compared with a randomly organized set of component buttons, well-organized functional grouping resulted in shorter scanpaths, covering smaller areas. The poorer interface resulted in more, but similar duration, fixations than the better interface. Whereas the poor interface produced less efficient search behavior, the layout of component representations did not influence their interpretability. Overall, data obtained from eye movements can significantly enhance the observation of users' strategies while using computer interfaces, which can subsequently improve the precision of computer interface evaluations.Relevance to industryThe software development industry requires improved methods for the objective analysis and design of software interfaces. This study provides a foundation for using eye movement analysis as part of an objective evaluation tool for many phases of interface analysis. The present approach is instructional in its definition of eye movement-based measures, and is evaluative with respect to the utility of these measures.
Book
Reading is a highly complex skill that is prerequisite to success in many societies in which a great deal of information is communicated in written form. Since the 1970s, much has been learned about the reading process from research by cognitive psychologists. This book summarizes that important work and puts it into a coherent framework. Note that the full-text of this book is not available.
Article
The purpose of this study was to explore how users perceive the affordances of an object from specific operation tasks. These affordances can be specified in terms of perceptual information. To this end, the eye-tracking technique was employed to capture what product parts uses' fixation focused on when using a product for specific operational tasks. The results show that, for physical-behavioral tasks, users were more interested in the physical properties that presented the affordance for product parts' operation-ability and assembly-ability; for functional tasks, users were more interested in the artificial signs that presented the affordance for product parts' functionality. In addition, this study proposed a feasible approach to exploring how users perceive affordances, which is contributive to applying the concept of affordance in product design.