How Healthcare Professionals Comprehend Process
Models - An Empirical Eye Tracking Analysis
Michael Winter1, Cynthia Bredemeyer2, Manfred Reichert1, Heiko Neumann3, Thomas Probst4, R¨
udiger Pryss2
1Institute of Databases and Information Systems, Ulm University, Germany
2Institute of Clinical Epidemiology and Biometry, University of W¨
urzburg, Germany
3Institute of Neural Information Processing, Ulm University, Germany
4Department for Psychotherapy and Biopsychosocial Health, Danube University Krems, Austria
{michael.winter, manfred.reichert, heiko.neumann}@uni-ulm.de, ruediger.pryss@uni-wuerzburg.de,
Abstract—Digitization is advancing rapidly in many preva-
lently analogue domains such as healthcare. For the latter
domain, the synergies with modern information technologies (IT)
have become an integral part regarding communication and
collaboration. For this reason, a comprehensible language is of
importance in order to allow a frictionless exchange of informa-
tion between domain experts. The Business Process Model and
Notation (BPMN) 2.0 represents a promising notation that may be
applied as lingua franca. Although the BPMN 2.0 is widespread
applied by experts in business and industry, little experience
exists how BPMN 2.0 is adopted in healthcare. In order to
assess how BPMN 2.0 is deployed in healthcare, we conducted
a preliminary eye tracking study, in which n = 16 professionals
from healthcare comprehended a particular BPMN 2.0 process
model. The results indicate that BPMN 2.0 might be a candidate
for a lingua franca to foster the comprehensible exchange of
information as well as collaboration between healthcare and IT.
Index Terms—Business Process Model and Notation 2.0, Pro-
cess Model Comprehension, Study, Eye Tracking, Healthcare
I. INTRODUCTION
Technological advancements in the context of digitization
are rapid and they are changing the way we work, com-
municate, and collaborate. Digitization no longer solely af-
fects the traditional information technology (IT) domain, but
domains across all industries such as economy, education,
and healthcare [1]. Regarding the latter, the ever-increasing
linkage of IT with the analogue healthcare world leads to new
opportunities (e.g., patients receive optimized treatment plans
[2] or operations can be performed remotely [3]). However, not
only opportunities emerge, but new challenges arise that have
to be thoroughly addressed [4]. In order to cope with these
challenges in healthcare, it is of utmost importance to provide
a language, which allows for the comprehensible exchange
of information between the IT and healthcare professionals
(i.e., lingua franca). A best practice in this context is the use
of visual process models [5], which conceptualize, determine,
and describe procedures, technical systems, or the processes of
organizations for the purpose of proper documentation, instant
communication, effective analysis, and collaboration [6].
In order to benefit from the merits of process models, a
proper comprehension of such models must be ensured [7].
Therefore, it is important to understand how individuals read
and comprehend process models, including positive and neg-
ative factors that affect model comprehension. If factors that
incapacitate process model comprehension are not addressed
properly, respective processes might not deliver the required
results. Prior research demonstrated that not only process
model expertise is a decisive factor, but process model com-
prehension also depends on other person-related characteristics
[8]. In this context, the utilization of eye tracking has become
a valuable instrument to investigate comprehension strategies,
difficulties to comprehend model constructs, or the cognitive
load during the comprehension of such models [9].
This paper presents the results from an exploratory eye
tracking study about how healthcare professionals read and
comprehend process models. Participants from healthcare
needed to comprehend a BPMN 2.0 process model, for which
we relied on eye tracking to record eye movements during
model comprehension. An emphasis was put on person-related
characteristics (e.g., gender, age) and their influence on the
comprehension of such models. The results indicate that the
presented process model could be comprehended intuitively,
but that there were differences in comprehension based on
several characteristics. The insights provide further informa-
tion towards a lingua franca to improve communication and
collaboration between IT and healthcare professionals.
Related work is presented in Section II. The study setting
is described in Section III, while Section IV discusses the
obtained results. Finally, Section V concludes the paper.
II. RELATED WORK
Since process models are becoming increasingly relevant
in digitization, a vast body of research has evolved in the
context of process model comprehension over the last decade.
Research studied objective as well as subjective comprehen-
sibility and the inherent factors with their influence on the
comprehension of such models. Regarding objective compre-
hensibility, structural properties of a process models have been
considered [7], [10]. In the context of subjective comprehen-
sibility, process model reader related factors such as model
expertise [11] or model reader preferences [12] were investi-
gated. As known from other domains, emphasis is increasingly
put on cognitive aspects. For example, cognitive style [13]
TABLE I
DEMOGRAPHIC DATA
ID Gender Age Prof. Exp. Tech. Prof. Tech. Lei. ID Gender Age Prof. Exp. Tech. Prof Tech. Lei.
1 female 26 - 35 Epi. No No No 9 female 26 - 35 Eco. Yes No No
2 male 26 - 35 Eco. Yes No No 10 female 20 - 25 Eco. Yes No No
3 male 20 - 25 Eco. No No No 11 male 26 - 35 Com. Yes Yes Yes
4 male 26 - 35 Com. Yes Yes Yes 12 female 20 - 25 Bio. No Yes No
5 female 20 - 25 Com. Yes Yes No 13 male 26 - 35 Com. Yes Yes Yes
6 male 26 - 35 Med. No Yes No 14 male 26 - 35 Med. Yes Yes No
7 female 26 - 35 Med. No No No 15 male 26 - 35 Med. No No No
8 female 20 - 25 Bio. No Yes No 16 female 20 - 25 Psy. No No No
or cognitive bias [14]. Furthermore, many technology-driven
approaches like the use of electrodermal activity [15] are
performed for a more fine-grained investigation of cognitive
aspects and their influence during process model comprehen-
sion. Especially the application of eye tracking is becoming
prominent for research on process model comprehension (e.g.,
[16]–[18]). Finally, the isolated consideration of specified
collectives during process model comprehension has revealed
various new insights [19]. However, to understand how health-
care professionals read and comprehend process models to
bridge the gap between IT and healthcare still needs more
investigations like shown in this paper.
III. STUDY SETTING
Healthcare professionals face a plethora of challenges when
performing their processes everyday [20]. Through the digiti-
zation, it is becoming increasingly apparent that processes,
communication, and medical applications in healthcare are
inadequately supported by the IT [21]. Therefore, it is im-
portant to foster a lingua franca in order to effectively fa-
cilitate the adoption of digitization and to benefit from the
associated advantages. One promising approach constitutes the
application of process models, which can be utilized for the
communication of information as well as collaboration. For an
efficient use of process models, it must be ensured that they
are comprehended correctly. Prior research showed that pro-
cess models can be comprehended intuitively, but especially
those without any process model expertise show difficulties in
coping with the complexities involved in parsing the relevant
information in such models [8]. Since IT and healthcare
coalesce in digitization, it is mandatory to investigate how
healthcare professionals read and comprehend process models.
Therefore, this study was carried out as part of a healthcare
research project by a student of translational medicine (with
dental background). The study had the goal to assess the
comprehension of a process model by healthcare professionals
to evaluate the suitability of BPMN 2.0 as lingua franca for
the communication between IT and healthcare professionals.
A. Study Planning
Participants: The study included n = 16 participants from
different disciplines in healthcare (i.e., 1 epidemiologist, 4
economists, 4 computer scientists, 4 medical practitioners, 2
biologists, and 1 psychologist; see Table I). Emphasis was
put on the following person-related characteristics and their
influence on the comprehension of process models: gender,
age, profession, process model expertise, and technology
affinity in profession as well as leisure. Based on these
person-related characteristics, the participants were divided
into different groups, which shared the same characteristics.
Note that participants could belong to several groups.
Materials: In this study, a BPMN 2.0 process model was
used [22]. The choice to use a BPMN 2.0 process model was
made for several rationales: BPMN 2.0 is the de facto industry
standard for the creation of comprehensible process models,
plus it is an ISO/IEC 1950:2013 norm. During the last decade,
a vast body of knowledge evolved, which has promoted the
widespread application of BPMN 2.0. The used process model
was composed of basic as well as advanced modeling elements
and contained a total number of 41 elements (excluding the
edges). Thematically, it documented a dental appointment with
all relevant participants, functions, and resources (see Fig. 1).
Thereby, particular emphasis was put in the comprehension
of the modeling constructs decision, loop, and parallelism, as
these constructs usually require modeling expertise.
Performance Measures:
1) Number of fixations: Fixations constitute eye movements
to deploy over attention by analyzing content (e.g., images)
with high resolution spectral analysis [23]. Further, fixation se-
quence analyses (i.e., sequential analysis of attentional content
reading) allowed us to make conclusion about the cognitive
load as well as about specific points in the stimulus that may
pose a challenge in the comprehension process.
Fig. 1. BPMN 2.0 process model with highlighted modeling constructs
decision, loop, and parallelism
TABLE II
DESCRIPTIVE RESULTS
ID Fix. Fix. Dur. Score ICL ECL GCL PUU PEU ID Fix. Fix. Dur. Score ICL ECL GCL PUU PEU
1 —– —– 7 3.5 2 6 17 17 9 605 211.57 7 3.5 2 5.5 16 15
2 528 226.61 7 2 1 7 15 18 10 704 179.13 6 3 2.6 4.5 16 15
3 594 218.34 7 2 3 6 11 16 11 610 264.69 7 3 2 6 18 16
4 —– —– 5 1.5 1.3 6 12 15 12 757 226.01 5 5 7 7 23 15
5 723 232.54 3 3.5 1.3 7 17 17 13 539 201.43 4 3.5 3.6 7 17 16
6 —– —– 5 5.5 2.6 5.5 16 16 14 591 311.95 7 3.5 2.3 6 14 14
7 631 213.97 6 4 3 4.5 14 13 15 753 163.36 6 4 5.6 6 12 11
8 677 185.32 7 5 4.3 7 16 15 16 767 231.32 5 4 3 6.5 19 18
All 652.53 220.46 5.87 3.53 2.93 6.10 15.43 15.81
(79.42) (36.53) (1.22) (1.06) (1.55) (0.79) (1.73) (2.83)
2) Fixation duration: The fixation duration indicates the
period of time in which the eyes remain still while looking
at a stimulus [24]. During this time period, the acquisition of
information from the viewed point in a stimulus (i.e., process
model) takes place. Hence, the analysis of the average fixation
duration allowed us for additional assumptions regarding the
cognitive load during process model comprehension.
3) Score: Participants needed to answer eight true-or-false
comprehension questions about the presented process models.
The comprehension questions referred to the semantic and
syntactic dimensions of the process model.
4) Cognitive load: The cognitive load depicts the invested
cognitive capacity of the working memory during a task and
consists of the following dimensions: intrinsic, extraneous,
and germane cognitive load [25]. Intrinsic load constitutes the
complexity of intrinsic information and is affected by existing
knowledge and element interactivity (e.g., demand on the
working memory). In turn, extraneous load is affected by the
way information is presented. Finally, germane load describes
the mental effort to process and comprehend information
based on constructed mental models. he adapted measurement
proposed in [26] was used for the study to measure the
dimensions. The single dimensions, which were comprised of
several items (i.e., two for intrinsic, three for extraneous, and
germane cognitive load), were assessed on a psychometrically
standardized questionnaire (i.e., 7-point Likert scale).
5) Perceived usefulness for understandability (PUU): It
describes the perceived usefulness of the process model and
is a part of the technology acceptance model (TAM) [27].
Therefore, four items on a 7-point Likert scale from strongly
disagree (i.e., 1) to strongly agree (i.e., 7) needed to be
answered, resulting in a min/max value of (4 x 7) [28].
6) Perceived ease of understandability (PEU): Derived
from TAM, PEU characterizes that the application of process
models is associated with less mental effort. Therefore, four
items on a 7-point Likert scale from strongly disagree (i.e., 1)
to strongly agree (i.e., 7) needed to be answered, resulting in
a min/max value of (4 x 7) [28].
B. Study Design and Procedure
The study was conducted in a designated lab at the Uni-
versity of W¨
urzburg. Due to the coronavirus (COVID-19)
pandemic, a dedicated study procedure was carried out in
compliance with hygiene regulations. Therefore, only one
participant could be evaluated at each study session. A session
took about 20 minutes and was as follows: The participant was
welcomed and first of all, he or she needed to sign a COVID-
19 declaration. Afterwards, the study procedure was explained
and an informed consent from the participant was obtained.
Then, the participant was asked to answer a set of demographic
questions for collecting relevant person-related characteristics
(e.g., age or gender). Following this, the used eye tracker was
calibrated with a 13-point screen-based calibration to ensure a
precise recording of the eye movements. After these steps, the
dental appointment BPMN 2.0 process model was presented
to the participant for exactly 03:30 minutes. During this time,
the participant had to read and comprehend the process model,
whilst their eye movements were recorded. The presentation
time was defined by three process modeling experts in a
consensus-building process, which has been deemed sufficient
to read and comprehend all relevant information in the process
model. After the comprehension task, the participant needed
to answer a questionnaire capturing the cognitive load and the
level of acceptability. Then, a comprehension task with eight
questions regarding the prior comprehended process model
was handed over in order to check the comprehensibility.
Finally, after the opportunity to leave feedback, the study
ended.
Instrumentation: COVID-19 declaration, demographic
data, comprehension question score, and information related to
the cognitive load as well as the level of acceptability was col-
lected with paper-based questionnaires. Eye movements were
recorded with the SMI iView X Hi-Speed system (sampling
rate of 240 Hz). The tracking appliance was placed in front of
a 23” monitor (resolution of 1920x1080, 96 PPI) showing the
process model to the participants. For calibration, a 13-point
calibration was performed. Eye tracking data collected during
the study was analyzed and visualized with SMI BeGaze 3.7.
Finally, SPSS 25 was used for all statistical analyses.
IV. DATA ANALYSIS AND INTERPRETATION
In Table II, for each participant (i.e., ID), the number of
fixations, average fixation duration (in ms), comprehension
question score, intrinsic, extraneous, germane cognitive load,
perceived usefulness for understandability, and perceived ease
of understandability are shown. Note that for participants 1, 4,
(a) Scan path (i.e., sequence of fixations) from a participant (b) Heat Map (i.e., density of fixations after model comprehension)
Fig. 2. Visualization of recorded eye movements
and 6, no eye tracking data are available because recording was
not possible due to issues with the eye movement calibration
(e.g., strabism). The mean and standard deviation of all
participants in all performance measures are presented as well.
The participants fixated in average 652.53 points (e.g.,
modeling constructs) during model comprehension. The av-
erage fixation duration alluded a medium level of interaction
with the model elements during comprehension (i.e., fixation
duration typically ranges from 150 to 300 ms). Participants
achieved an average score of 5.87 (max. is 8). This indicates
that the presented dental appointment process model could be
comprehended intuitively. Concerning cognitive load, while
intrinsic (i.e., 3.53) and extraneous (i.e., 2.93) load were at
a moderate level, the result for germane cognitive load (i.e.,
5.87) indicated an above-average level. Hence, the mental
effort for comprehending the information in the model was
supported by constructed mental schemata, which led to a
more proper comprehension. Regarding level of acceptability,
the results for PUU (i.e., 15.43) and PEU (i.e., 15.81) were
on average (i.e., max. is 28 respectively), which indicated
that the participants were indecisive regarding the benefits of
presenting process information in a BPMN process model.
Regarding eye movements, scan paths (i.e., chronological
alignment of fixation sequence) during model comprehension
with related properties (e.g., fixation) were considered. Fig. 2
(a) depicts a scan path from a participant. The dots represent
fixations (i.e., the larger the dot, the longer the dwell time)
and the connecting lines constitute eye movements from one
fixation to another. The participant has looked at each element
in the process model and it is recognizable by the fixations,
which places in the model have been looked at more often
and longer. It is noticeable that the constructs decision, loop,
and parallelism have been considered frequently compared to
other elements. Fig. 2 (b) shows the superimposed heat map
of all participants. A heat map provides no information about
individual fixations, in lieu the magnitude of visual attention
for all participants is revealed. The heat map shown in Fig 2
(b) confirmed similar eye movements of all participants, as the
eye movements of the participant in Fig. 2 (a). 1.
1Materials and results are available at: tinyurl.com/1uops3x0
A. Inferential Statistics
To evaluate whether person-related characteristics (see
Tab. I) have an influence on process model comprehension,
the participants were grouped based on the same characteris-
tics. For each person-related characteristic, a two-independent
Mann-Whitney U test was performed for each performance
measure (see Sect. II). For the characteristic Profession, a one-
way analysis of variance (ANOVA) was performed. Finally,
the significance value was set to p <.05.
1) Gender: The results of 8 female participants were
compared with 8 males. Regarding the performance measure
Fixation and Fixation Duration, the results of 7 female
participants were compared with 6 males.
Score: U = 29.0, z = -.33; p = .739
ICL: U = 18.5, z = -1.45; p = .148
ECL: U = 27.0, z = -.53; p = .597
GCL: U = 31.0, z = -.11; p = .913
PUU: U = 31.5, z = -.05; p = .739
PEU: U = 15.0, z = -1.81; p = .071
Fixation: U = 6.0, z = -2.14; p = .032
Fix. Dur.: U = 17.0, z = -.57; p = .568
The U test indicated that the number of fixations was greater
for female (Median (Mdn) = 704.0) than for male participants
(Mdn = 566.5).
2) Age: The results of 6 participants with an age between
20 - 25 were compared with 10 with an age between 26 - 35.
Regarding the performance measure Fixation and Fixation
Duration, the results of 6 participants with an age between
20 - 25 were compared with 7 with an age between 26 - 35.
Score: U = 23.0, z = -.80; p = .423
ICL: U = 25.0, z = -.55; p = .580
ECL: U = 18.0, z = -1.31; p = .190
GCL: U = 19.0, z = -1.25; p = .213
PUU: U = 23.5, z = -.72; p = .470
PEU: U = 20.0, z = -1.10; p = .272
Fixation: U = 7.0, z = -2.00; p = .046
Fix. Dur.: U = 20.0, z = -.14; P = .886
The U test indicated that the number of fixations was greater
for participants with an age between 20 - 25 (Mdn = 713.5)
than for an age between 26 - 35 (Mdn = 605.0).
3) Profession: The results of 1 epidemiologist, 4
economists, 4 computer scientists, 4 medical practitioners, 2
biologists, and 1 psychologist were compared with each other.
Therefore, the performance measure Fixation and Fixation
Duration, the results of 4 economists, 3 computer scientists,
3 medical practitioners, 2 biologists, and 1 psychologist were
compared with each other.
Score: F(5,10) = 1.52; p = .268
ICL: F(5,10) = 2.51; p = .053
ECL: F(5,10) = 2.67; p = .088
GCL: F(5,10) = 1.52; p = .269
PUU: F(5,10) = 2.57; p = .096
PEU: F(5,10) = 1.65; p = .234
Fixation: F(4,12) = .646; p = .645
Fix. Dur.: F(4,12) = .233; p = .912
No significant differences were found.
4) Expertise: The results of 8 participants with process
model expertise were compared with 8 without expertise.
Regarding the performance measure Fixation and Fixation
Duration, the results of 6 participants with process model
expertise were compared with 7 without expertise.
Score: U = 31.5, z = -.06; p = .956
ICL: U = 8.5, z = -2.52; p = .012
ECL: U = 8.5, z = -2.49; p = .013
GCL: U = 30.5, z = -.17; p = .869
PUU: U = 32.0, z = .00; p = 1.00
PEU: U = 29.0, z = -.32; p = .747
Fixation: U = 8.0, z = -1.86; p = .063
Fix. Dur.: U = 14.0, z = -1.00; p = .317
The U test indicated that the intrinsic cognitive load was
greater for participants without process model expertise (Mdn
= 4) than for participants with expertise (Mdn = 3.25). The U
test indicated that the extrinsic cognitive load was greater for
participants without process model expertise (Mdn = 3) than
for participants with expertise (Mdn = 2).
5) Technical Affinity in Profession: The results of 8
participants without technical affinity in their profession
were compared with 8 with technical affinity. Regarding the
performance measure Fixation and Fixation Duration, the
results of 7 participants without technical affinity in their
profession were compared with 6 with a technical affinity.
Score: U = 19.5, z = -1.39; p = .166
ICL: U = 25.5, z = -.69; p = .486
ECL: U = 31.5, z = -.05; p = .958
GCL: U = 18.0, z = -1.54; p = .125
PUU: U = 30.0, z = -.22; p = .830
PEU: U = 22.0, z = -1.06; p = .288
Fixation: U = 19.0, z = -.29; p = .775
Fix. Dur.: U = 12.0, z = -1.29; p = .199
No significant differences were found.
6) Technical Affinity in Leisure: The results of 13
participants without a technical affinity in their leisure were
compared with those of 3 with a technical affinity in their
leisure. Regarding the performance measure Fixation and
Fixation Duration, the results of 11 participants without a
technical affinity in their leisure were compared with those
of 2 with a technical affinity in their leisure.
Score: U = 13.5, z = -.85; p = .394
ICL: U = 7.5, z = -1.65; p = .099
ECL: U = 14.5, z = -.68; p = .498
GCL: U = 16.0, z = -.49; p = .623
PUU: U = 18.0, z = -.21; p = .836
PEU: U = 16.5, z = -.41; p = .683
Fixation: U = 6.0, z = -.99; p = .324
Fix. Dur.: U = 9.0, z = -.40; p = .693
No significant differences were found.
B. Discussion and Next Steps
The results revealed that the comprehension of the process
model was intuitively possible for healthcare professionals.
Although individual results were significant, no person-related
characteristic has been identified, which considerably favors
process model comprehension. Expertise in process modeling
assisted the working memory load during model comprehen-
sion. The summarized results of all participants from the
study indicated an average interactivity of the process model
elements and the presented process model was positively
perceived. Particularly interesting was the above-average ger-
mane cognitive load. It appears that participants were not
strongly mentally engaged and, hence, were not confronted
with severe difficulties in handling the information shown
in the process model. Regarding the level of acceptability,
the results were on an average level, which implies that
the behavioral intention to apply process model is not clear.
Additionally, participants were indecisive in their attitude
regarding the benefits of process models as a lingua franca
for communication and collaboration. Regarding recorded eye
movements, the descriptives show that similar comprehension
strategies were applied, whereas the model constructs decision,
loop, and parallelism have been considered most often by
the participants. However, several observations pertaining to
the eye movements showed that participants of older age
needed fewer fixations and achieved the same comprehension
question score juxtaposed to participants of younger age.
This result may indicate person-related characteristics, which
are influencing the nature of comprehending process models.
In this study, we identified the characteristics gender, age,
and expertise as such influencing factors that needed to be
investigated in further studies (see Section IV-C). Importantly,
participants without expertise in process modeling were able
to comprehend the syntactic of the respective constructs. An
explanation would be that the consideration of the semantics
supports the derivation of the syntactics. The insights obtained
from this study unraveled that BPMN process models might
be a suitable lingua franca to foster the comprehensible
exchange of information as well as collaboration between IT
and healthcare professionals in the course of digitization. In
order to get a better understanding, more studies and results
are required. Therefore, the exploratory insights from this
study can be used as connecting factor. These include, for
example, the comprehension of more complex process models
documenting different scenarios. In addition, the application
of a latent variable model might reveal specific person-related
characteristics, which can have a significant positive and a
negative influence on the comprehension of such models.
Attention should be paid on measures that increase the level
of acceptability regarding the proper utilization and the related
benefits in applying process models. Based on the preliminary
results, we are planning a follow-up study. We will consider
additional person-related characteristics, eye movements, and
the electrodermal activity that may provide novel insights.
C. Limiting Factors
First, the process model might not be representative. Process
models document procedures of the real world, which are more
complex. However, the used process model was kept simple
intentionally. Research showed that process model compre-
hension becomes more error-prone with increasing number
of modeling elements [29]. Consequently, complex process
models make different demands regarding the cognitive load
and the level of acceptability. Second, the scenario documented
in respective process model represents another limitation. Only
one scenario (i.e., dental appointment) was presented and,
hence, more different scenarios in process models might have a
diverse influence on model comprehension. Third, the inherent
difficulty (e.g., question difficulty) of the study material may
not be appropriate. In detail, the true-or-false comprehension
questions might be too easy, since the participants reached
an above-average score. Fourth, another limitation lies in
the recruitment of participants. Regarding several person-
related characteristics (e.g., profession, age), an imbalance
was present, which could have affected the statistics. Fifth,
the sample sizes limit the statistical power and there might
be significant differences, which we could not detect, but
which might become apparent in larger sample sizes. Finally,
sixth, current COVID-19 hygiene regulations (e.g., permanent
wearing of a face mask) may have influenced the participants.
V. SUMMARY
In the context of digitization, this paper presented the results
from an exploratory eye tracking study on process model
comprehension. The objective of this study was to contribute
additional insights for providing a lingua franca to emerge new
synergies between IT and healthcare professionals in terms of
communication as well as collaboration. Professionals from
healthcare were asked to read and comprehend a BPMN 2.0
process model, whilst their eye movements as well as infor-
mation regarding the cognitive load and level of acceptability
were captured. The results revealed that the process model can
be comprehended intuitively and that BPMN 2.0 might be a
candidate for a lingua franca in this context. While results look
promising, however, further studies are needed as the study at
hand was confronted several limitations (e.g., one involving
more participants with balanced distribution of person-related
characteristics as well as more diverse process models). The
results foster our conceptual framework for process model
comprehension in healthcare that incorporates methods and
theories from cognitive neuroscience and psychology [30].
ACKNOWLEDGMENT
The study was carried out in the context of the translational
medicine graduate program of the University of W¨
urzburg.
REFERENCES
[1] M. M. Gobble, “Digitalization, digitization, and innovation,” Research-
Technology Management, vol. 61, no. 4, pp. 56–59, 2018.
[2] M. H. van Velthoven, C. Cordon, and G. Challagalla, “Digitization of
healthcare organizations: the digital health landscape and information
theory,” Int J Medical Informatics, vol. 124, pp. 49–57, 2019.
[3] H. Sadeghian, F. Zokaei, and S. H. Jazi, “Constrained kinematic control
in minimally invasive robotic surgery subject to remote center of motion
constraint,” Journal of Int & Robo Sys, vol. 95, no. 3, pp. 901–913, 2019.
[4] M. Wehde, “Healthcare 4.0,” IEEE Engineering Management Review,
vol. 47, no. 3, pp. 24–28, 2019.
[5] G. Adamo et al., “What is a process model composed of?” Soft and Sys
Mod, pp. 1–29, 2021.
[6] D. C. Wynn and P. J. Clarkson, “Process models in design and devel-
opment,” Res in Eng Des, vol. 29, no. 2, pp. 161–202, 2018.
[7] K. Figl, “Comprehension of procedural visual business process models,”
Bus & Inf Sys Eng, vol. 59, no. 1, pp. 41–67, 2017.
[8] M. Zimoch et al., “Cognitive insights into business process model
comprehension: Preliminary results for experienced and inexperienced
individuals,” in BPMDS’17, 2017, pp. 137–152.
[9] ——, “Eye tracking experiments on process model comprehension:
Lessons learned,” in BPMDS’17, 2017, pp. 153–168.
[10] O. Turetken et al., “The effect of modularity representation and presen-
tation medium on the understandability of business process models in
bpmn,” in BPM’16. Springer, 2016, pp. 289–307.
[11] J. Claes et al., “The structured process modeling method (spmm) what
is the best way for me to construct a process model?” Decision Support
Systems, vol. 100, pp. 57–76, 2017.
[12] K. Figl and J. Recker, “Exploring Cognitive Style and Task-Specific
Preferences for Process Representations,” in Req Eng, 2014, pp. 63–85.
[13] O. Turetken and Others, “Cognitive Style and Business Process Model
Understanding,” in CAiSE’17, 2017, pp. 72–84.
[14] M. Razavian et al., “When cognitive biases lead to business process
management issues,” in BPM’16, 2016, pp. 147–156.
[15] M. Winter et al., “Towards the applicability of measuring the electroder-
mal activity in the context of process model comprehension: Feasibility
study,” Sensors, vol. 20, no. 16, p. 4561, 2020.
[16] R. Petrusel et al., “How visual cognition influences process model
comprehension,” Dec Supp Sys, vol. 96, pp. 1 – 16, 2017.
[17] W. Wang et al., “Effect of Linked Rules on Business Process Model
Understanding,” in BPM’17, 2017, pp. 200–215.
[18] R. B. Duarte et al., “A systematic literature review on the usage of
eye-tracking in understanding process models,” BPMJ, 2020.
[19] M. Tallon et al., “Comprehension of business process models: Insight
into cognitive strategies via eye tracking,” ESWA, pp. 145–158, 2019.
[20] M. Ramos-Merino et al., “A bpmn based notation for the representation
of workflows in hospital protocols,” J of MS, vol. 42, pp. 1–10, 2018.
[21] P. Carayon et al., “Challenges and opportunities for improving patient
safety through human factors and systems engineering,” Health Affairs,
vol. 37, no. 11, pp. 1862–1869, 2018.
[22] OMG, “Object Management Group Specification. Business Process
Modeling & Notation 2.0,” 2019.
[23] I. Krajbich, C. Armel, and A. Rangel, “Visual fixations and the compu-
tation and comparison of value in simple choice,” Nature Neuroscience,
vol. 13, no. 10, pp. 1292–1298, 2010.
[24] J. M. Henderson, W. Choi, S. G. Luke, and J. Schmidt, “Neural correlates
of individual differences in fixation duration during natural reading,”
Quart J of Experimental Psychology, vol. 71, no. 1, pp. 314–323, 2018.
[25] F. Paas et al., “Cognitive load theory and instructional design: Recent
developments,” Edu Psych, vol. 38, no. 1, pp. 1–4, 2003.
[26] M. Klepsch, F. Schmitz, and T. Seufert, “Development and validation of
two instruments measuring intrinsic, extraneous, and germane cognitive
load,” Frontiers in Psychology, vol. 8, Nov. 2017.
[27] W. R. King and J. He, “A meta-analysis of the technology acceptance
model,” Inf & Man, vol. 43, no. 6, pp. 740–755, Sept. 2006.
[28] O. Turetken et al., “Influential characteristics of enterprise information
system user interfaces,” J of Comp Inf Sys, vol. 59, pp. 243–255, 2019.
[29] J. Mendling et al., “Understanding the occurrence of errors in process
models based on metrics,” in CoopIS’07, 2007, pp. 113–130.
[30] M. Zimoch et al., “Towards a Conceptual Framework Fostering Process
Comprehension in Healthcare,” in CBMS’17, 2017, pp. 167–168.
