Considering Social Distance as an Influence
Factor in the Process of Process Modeling
Michael Zimoch, Jens Kolb, and Manfred Reichert
Institute of Databases and Information Systems, Ulm University, Germany
{michael.zimoch,jens.kolb,manfred.reichert}@uni-ulm.de
Abstract. Enterprise repositories comprise numerous business process
models either created by in-house domain experts or external business
analysts. To enable a widespread use of these process models, high model
quality (e.g., soundness) as well as a sufficient level of granularity are
crucial. Moreover, they shall reflect the actual business processes prop-
erly. Existing modeling guidelines target at creating correct and sound
process models, whereas there is only little work dealing with cognitive
issues influencing model creation by process designers. This paper ad-
dresses this gap and presents a controlled experiment investigating the
construal level theory in the context of process modeling. In particu-
lar, we investigate the influence the social distance of a process designer
to the modeled domain has on the creation of process models. For this
purpose, we adopt and apply a gamification approach, which enables
us to show significant differences between low and high social distance
with respect to the quality, granularity, and structure of the created pro-
cess models. The results obtained give insights into how enterprises shall
compose teams for creating and evolving process models.
1 Introduction
Due to the increasing adoption of process-aware information systems (PAIS),
contemporary enterprise repositories comprise large collections of process mod-
els [1]. Usually, process models vary in respect to their quality and level of
granularity. Further, they face a wide range of problems affecting model under-
standability and error probability [2]. However, high quality of process models is
crucial for enterprises to guarantee proper process implementation and execution
in a PAIS [3]. As a prerequisite, process models should reflect the actual busi-
ness processes properly and at the right level of granularity [4]. To address this
issue, considerable work on criteria related to process model quality and com-
prehensibility has been conducted [5, 6]. In addition, modeling guidelines exist
that support process designers in creating process models of high quality [7, 8].
There is only little work evaluating the influence of cognitive aspects on the
process of process modeling [9] as well as their effects on the resulting process
models [10, 11]. If we do not understand these cognitive aspects, however, pro-
cess modeling projects might not deliver proper artifacts or even fail. This paper
2 Michael Zimoch, Jens Kolb, and Manfred Reichert
investigates a fundamental factor presumably influencing the process of process
modeling, i.e., social distance [12]. The latter is a well-established notion in the
Construal Level Theory (CLT), constituting an important part of psychological
distance [13]. In this context, studies have shown that human thinking and acting
are both strongly influenced by psychological distance [14]. According to CLT,
we experience only the here and now, and form an abstract mental construal of
distant objects or events [12, 14]. For example, when attending a music festival,
one is able to undergo the whole festival atmosphere. In turn, watching the fes-
tival on television, the focus is more on the line-up and, hence, the performances
of the bands, i.e., experience is more superordinate.
Sect. 2 introduces CLT. Gamification and the considered process scenario
are described in Sect. 3. Sect. 4 introduces the research question addressed and
defines the experiment setting. Sect. 5 deals with experiment preparation and
its execution. Results are presented and analyzed in Sect. 6. Finally, Sect. 7
discusses related work and Sect. 8 summarizes the paper.
2 Background on Construal Level Theory
Construal Level Theory (CLT) describes the effects psychological distance has
on objects or events [12, 14]. Generally, CLT states that increasing psychological
distance affects our mental representation of these objects or events. In turn,
this influence on human perception has a strong impact on our actions and
thoughts [13]. The reason behind this phenomenon is the so-called level of con-
strual (LOC), which describes how individuals interpret and perceive objects
and events. Increasing psychological distance affects the cognitive abilities and
leads to a change in the perception of an object or event.
CLT describes two levels of thinking: low- and high-level construal.High-level
construals are abstract, decontextualized, coherent, and superordinate represen-
tations compared to low-level construals. If an object or event is further away,
we think about it in terms of high-level construals. However, the smaller the
distance to objects or events is, the more we think in low-level construals. More-
over, these two levels of construals are influenced by psychological distance. While
objective distance describes the quantitative spatial distance in the real world,
psychological distance describes our feelings, thoughts, and emotions in relation
to an object or event. In turn, the latter is considered as psychologically distant,
if it is not experienced physically. For this case, a mental representation must
be constructed.
Psychological distance can be further subdivided into social, spatial,temporal,
and hypothetical distance [13, 15–17]. Social distance, on which we focus in this
paper, describes our relation to other individuals or accrues for events not being
self-experienced (cf. Fig. 1); e.g., whether or not choosing a seat in a bus being
more distant from a particular individual is directly reflected by the latter [18].
In previous research we already addressed the first characteristic, i.e., the
relation to other individuals [19]. More precisely, results showed a significant
influence of social distance on the quality and level of granularity of created
process models. In accordance with CLT, process models created by process de-
Social Distance in the Process of Process Modeling 3
signers with a low social distance revealed a higher quality as well as granularity
compared to process models created by process designers with a higher social
distance. Furthermore, process designers were more self-confident about the pro-
cess models they had create. Hence, the latter characteristic, i.e., event which is
not self-experienced, is evaluated in this paper.
Low DistanceHigh Distance
abstract, imprecise
specific, precise Person
Fig. 1. Construal Level Theory - Social Distance
3 Gamification, Virtual World, and 3D Scenario
In order to simulate variability with respect to social distance, a gamification
approach is applied, i.e., the benefits of gamification in a virtual world are used
in the context of process modeling. First, this allows for an adequate reflection
of the real world problem. Second, the motivation of subjects (i.e., participants
of an experiment) may increase. Third, an occurrence of the effects of social
distance may be ensured.
Gamification is the technique of using game elements, designs, and thinkings
in a non-game context to engage and motivate employees [20]; e.g., achievements
known from computer games are interpreted in enterprise software. As a con-
sequence, work becomes more enjoyable, thus resulting in higher efficiency [21].
Moreover, a virtual world constitutes a computer-simulated environment, using
the metaphor of the real world, but without its physical limitations [22]. In a
virtual world, individuals act as textual, 2D, or 3D avatars, i.e., as a controllable
proxy in the virtual world. Thus, they experience a degree of telepresence, i.e.,
an experience of presence in a remote location [23].
In the context of our experiment, relative to a real-world process from a
manufacturer of gardening tools, a process scenario related to the processing of
an order in a warehouse is contrived, which may be either experienced actively
or passively (cf. Sect. 4). The entire process takes place in a full 3D virtual
environment taking elements of gamification into account; e.g., exploring (i.e.,
learning more about the virtual construct) and puzzle elements (i.e., motivating
subjects to solve a problem). The 3D warehouse scenario is implemented with
Unity, a game development platform. In the realized scenario, subjects interact
with a 3D avatar using point and click game mechanics.
Following this, a description of the processing of an order in the warehouse is
provided. Fig. 2 shows the layout as well as the chronological progress through
the warehouse. The scenario starts in the office of the warehouse 1
. First, an
order is taken providing information on the items to be processed. Generally,
several items need to be processed by subjects in this context. At the storage
racks (cf. Fig. 3), subjects have the choice to get the items either with the
forklift or the picking system 2
. Since the forklift can carry only one pallet at
a time, the items must be collected sequentially. The picking system comprises
several grapplers that allow collecting all items either separately or at once.
4 Michael Zimoch, Jens Kolb, and Manfred Reichert
Then, items are disclosed at the collection point and checked for completeness
3
. Following this, the items need to be packed in appropriate boxes, which
are then palletized 4
. After placing each box on a pallet, subjects may decide
on how to transport the pallets to the shipping area, i.e., either by using the
forklift or the automatic loading system 5
. While the forklift can transport the
pallets only sequentially, the automatic loading system takes care of everything
automatically. As advantage of the automatic loading system, the subjects can
print the required delivery documents (i.e., bill of delivery and pallet receipts)
in parallel 6
. Thereafter, pallets are labeled with the printed pallet receipts and
are loaded on the trailer with the forklift 7
. Finally, the bill of delivery is placed
in the trailer and doors are closed.
Fig. 2. Layout of the Warehouse Scenario Fig. 3. Storage Racks
4 Research Question and Experiment Definition
This section introduces the definition and planning of the experiment for mea-
suring the influence of the social distance on the process of process modeling
and the resulting artifacts. Sect. 4.1 explains the context of the experiment and
defines its goal. Sect. 4.2 introduces the hypothesis considered for testing, and
Sect. 4.3 presents the experimental setup. Sect. 4.4 explains the design of the
experiment. Finally, Sect. 4.5 discusses factors threatening the validity of results.
4.1 Context Selection and Goal Definition
Business processes are either modeled by in-house process designers or external
ones. In this context, process designers are responsible for interviewing process
stakeholders and participants as well as for capturing the gathered knowledge
in process models. Usually, the process designers are not directly involved in
the processes to be modeled; e.g., they may be member of the quality assurance
department. In other cases, due to limited resources, enterprises assign such
modeling and analysis tasks to external resources; e.g., business analysts.
So far, it has not been well understood how an increased social distance
affects the quality, granularity, and structure of the resulting process models. To
close this gap, this paper investigates the following research question:
Is the process of process modeling, i.e., the quality, granularity, and structure of
the process models resulting from it, affected by the social distance the process
designers have on the respective business processes?
Social Distance in the Process of Process Modeling 5
Despite existing work on the quality [2, 8, 24, 25], granularity [26], and struc-
ture [27] of process models there is only little research addressing cognitive as-
pects of process modeling [10, 11, 28]. In particular, it is not well understood
whether certain cognitive aspects lead to minor process quality, i.e., deficiencies
regarding the pragmatic, semantic, perceived, and syntactic model quality.
Based on previous research (cf. [19]), this paper continues investigating the
influence social distance has on the process of process modeling and its outcomes.
As opposed to the previous experiment, where social distance was experienced by
the relation to other individuals, the presented experiment varies social distance
with a scenario (i.e., processing of an order in a warehouse) that may either
be experienced actively (i.e., low) or passively (i.e., high social distance) using
gamification. The goal can be formulated as:
Analyze process models
for the purpose of evaluating
with respect to their level of construal
from the point of view of the researchers
in the context of students and research staff.
4.2 Hypothesis Formulation
Based on the goal definition and taking CLT into account, six hypotheses are
derived. In detail, they investigate whether social distance influences the level of
construal during the process of process modeling or, more precisely, the quality,
granularity, and structure of the resulting process models:
Does the social distance influence the pragmatic quality when creating process models?
H0,1: There are no significant differences in the pragmatic quality when modeling processes with
low social distance compared to high social distance.
H1,1: There are significant differences in the pragmatic quality when modeling processes with low
social distance compared to high social distance.
Does the social distance influence the semantic quality when creating process models?
H0,2: There are no significant differences in the semantic quality when modeling processes with low
social distance compared to high social distance.
H1,2: There are significant differences in the semantic quality when modeling processes with low
social distance compared to high social distance.
Does the social distance influence the perceived quality when creating process models?
H0,3: There are no significant differences in the perceived quality when modeling processes with low
social distance compared to high social distance.
H1,3: There are significant differences in the perceived quality when modeling processes with low
social distance compared to high social distance.
Does the social distance influence the syntactic quality when creating process models?
H0,4: There are no significant differences in the syntactic quality when modeling processes with low
social distance compared to high social distance.
H1,4: There are significant differences in the syntactic quality when modeling processes with low
social distance compared to high social distance.
Does the social distance influence the level of granularity when creating process models?
H0,5: There are no significant differences in the level of granularity when modeling processes with
low social distance compared to high social distance.
H1,5: There are significant differences in the level of granularity when modeling processes with low
social distance compared to high social distance.
Does the social distance influence the process model structure when creating process models?
H0,6: There are no significant differences in the process model structure when modeling processes
with low social distance compared to high social distance.
H1,6: There are significant differences in the process model structure when modeling processes with
low social distance compared to high social distance.
6 Michael Zimoch, Jens Kolb, and Manfred Reichert
4.3 Experimental Setup
This section describes subjects,object, and response variables of the experiment
as well as its instrumentation and data collection procedure.
Subjects. Ideally, process designers are modeling experts. However, they usually
obtain only basic training and have limited process modeling skills [29]. From
subjects (i.e., students and staff members) we require that they are familiar with
process modeling although they were not experts in this area. A replication of
the experiment with modeling experts might lead to different results [30]. Hence,
results might not be generalizable for the entire population of process designers.
Object. The object is the outcome resulting from a stated modeling task, i.e.,
aprocess model expressed in terms of the Business Process Model and Nota-
tion (BPMN). To ensure familiarity of subjects with BPMN and to guarantee
that differences in response variables are not caused due to a lack of familiarity
with BPMN, but rather due to differences in social distance, we choose an easy
and understandable scenario. More precisely, the modeling task deals with the
processing of an order in a warehouse (cf. Sect. 3). Task descriptions are cre-
ated reflecting low and high social distance. One group is directly involved (i.e.,
low) in the process, while the other is only indirectly involved (i.e., high social
distance). For low social distance, subjects are actively playing the warehouse
scenario. In turn, regarding high social distance, subjects are watching the ware-
house scenario in a video. To ensure that there exist no interferences and there
is sufficient clearance between the two social distances, two pilot studies for each
social distance are performed. Respective task descriptions are kept rather ab-
stract to give subjects the possibility to model as detailed as they like.
Factor and factor levels. The factor considered in the experiment is social
distance with levels low and high social distance. Accordingly, the task descrip-
tion is adjusted to vary social distance, i.e., to model the order process either
after playing (i.e., low) or watching (i.e., high social distance) the scenario.
Response variable. As response variable, we consider the level of construal
that cannot be directly measured. Everything being distant from us is expressed
more abstractly (cf. Sect. 2). We assume that the level of construal impacts
the quality,granularity, and structure of the resulting process model. For this
purpose, process model quality is characterized by four dimensions, i.e., prag-
matic,semantic,perceived, and syntactic quality making use of semiotic theory,
i.e., SEQUAL framework [31, 32]. Pragmatic quality describes process model
comprehension. It is measured by the level of understanding. In turn, semantic
quality covers correctness, relevance, completeness, and authenticity of a process
model. Correctness expresses that all elements of a process model are correct.
Relevance signifies that all elements in the process model are relevant for the pro-
cess. Moreover, completeness implies that relevant aspects about the domain are
not missing, i.e., superfluous elements are considered as well. Finally, authenticity
expresses that the chosen representation gives a true impression of the domain.
Pragmatic quality and semantic quality are rated by two modeling experts in
a consensus-building process based on a 7-point Likert scale [33], i.e., from 0
(strongly disagree) to 6 (strongly agree). In turn, perceived quality depends on
Social Distance in the Process of Process Modeling 7
the degree to which a subject agrees with the resulting process model. It can be
subdivided into agreement,missing aspects,accurate description,mistakes, and
satisfaction [34]. Perceived quality is rated by each subject on a 5-point Likert
scale, ranging from 0 (strongly disagree) to 4 (strongly agree), after finishing the
modeling task. Agreement expresses to which degree the process model matches
with the actual business process. Missing aspects rates whether significant as-
pects are missing in the resulting process model. In turn, accurate description
expresses how accurately the process model matches the real world process.
Mistakes corresponds to the subject rate indicating whether there are serious
mistakes in the resulting process model. Finally, satisfaction expresses the de-
gree subjects are satisfied with the process models created by them. Syntactic
quality of a process model is measured by counting syntactical rule violations
of the applied modeling language, i.e., BPMN. Process granularity is measured
through the complexity of the resulting process models, i.e., simple metrics like
number of activities, gateways, nodes, edges, elements, and execution paths.Pro-
cess model structure is analyzed with the following process metrics: separability,
sequentiality, cyclicity, and diameter [3, 35]. Separability is defined as the ratio
of the number of cut-vertices to the total number of nodes in the process model.
Sequentiality, in turn, is the degree to which the process model is constructed of
pure sequences of tasks. Moreover, cyclicity relates to the number of nodes on
cycles to all nodes in the process model. Diameter gives the length of the longest
path from a start node to an end node in the process model. Fig. 4 summarizes
the response variables we consider in a research model.
4.4 Experimental Design
We apply guidelines for designing experiments as described in [36], and conduct
arandomized, balanced, and blocked single factor experiment. The experiment
is randomized since subjects are assigned to groups randomly and it is ensured
that both groups have same size (i.e., balanced). Moreover, subjects are grouped
(i.e., blocked) to not mix social distance. Finally, only a single factor varies, i.e.,
the level of construal. Fig. 5 illustrates this setup.
Instrumentation and data collection procedure. To precisely measure re-
sponse variables in a non-intrusive manner, we use the Cheetah Experimental
Platform (CEP) [9]. CEP provides a BPMN modeling environment that records
modeling steps and their attributes; e.g., timestamps and type of modeling ac-
tion. Resulting process models are then stored. Finally, demographic data and
qualitative feedback is gathered from subjects based on questionnaires.
4.5 Risk Analysis
Generally, any experiment bears risks that might affect its results. Thus, its va-
lidity or, more precisely, its levels of validity need to be checked, i.e., internal
validity (“Are effects caused by independent response variables?”) and external
validity (“May results be generalized?”).
Risks to internal validity.Risks that might influence the modeling outcome
8 Michael Zimoch, Jens Kolb, and Manfred Reichert
include process modeling experience of involved subjects and uneven distribu-
tions of subjects over two groups. Furthermore, post data validation ensures that
in both groups subjects are at least moderately familiar with process modeling
(cf. Sect. 5.3). It is assured that both groups show the same or similar familiar-
ity level, i.e., median is 3 for both groups on a 5-point Likert scale. Further, the
chosen modeling task constitutes a risk to internal validity. To ensure familiarity
of subjects and to guarantee that differences in quality, granularity, and struc-
ture are due to social distance, we choose an easy and comprehensible scenario
(cf. Sect. 3). To further ensure that subjects are not negatively influenced by
tiredness, boredom, or hunger, the experiment is conducted at a time of the day
for which the mentioned frame of mind can be excluded.
Risks to external validity.On one hand, the subjects have academic back-
ground (i.e., students and research staff), which might limit generalizability of
results. On the other, they rather have profound knowledge in process modeling
(cf. Sect. 5.3). We may consider them as proxies for professionals who have ob-
tained basic training so far. Further, process model quality may depend on the
appropriateness of the chosen modeling languages and tools. To mitigate this
risk, both groups use an intuitive process modeling tool as well as an established
modeling language (cf. Sect. 4.3). Finally, a potential risk for external validity
is that we measure social distance with one modeling task. To mitigate this and
to allow for generalizability, varying experiments need to be conducted.
Psychological Distance
F: Syntactic Quality
O: Number of Syntactical Errors
Quality
Legend:
F: Theoretical Factor
O: Factor Operationalization
F: Semantic Quality
O: Correctness
Relevance
Completeness
Authenticity
F: Perceived Quality
O: Agreement
Missing Aspects
Accurate Description
Mistakes
Satisfaction
Granularity
F: Granularity
O: Number of Activities
Number of Gateways
Number of Nodes
Number of Edges
Number of Elements
Number of Exec. Paths
F: Social Distance
O: Level of Social
Distance
F: Pragmatic Quality
O: Level of Understanding
Structure
F: Structure
O: Separability
Sequentiality
Cyclicity
Diameter
Fig. 4. Research Model
Low Social
Distance
(Playing)
Process Model
n Subjects
High Social
Distance
(Watching)
Process Model
Subject n/2+1
Task
.
.
..
.
.
Subject nSubject 1
Subject n/2
Subject
Group 1 Subject
Group 2
Factor
Object
Fig. 5. Experiment Design
5 Experiment Operation
Based on the provided experiment definition, Sect. 5.1 summarizes the experi-
ment preparation. Sect. 5.2 describes the execution of the experiment, and Sect.
5.3 deals with the validation of the data collected during the experiment.
5.1 Experiment Preparation
Students and research staff familiar with process modeling are invited to join the
experiment. Subjects are not informed about the aspects we want to investigate.
Social Distance in the Process of Process Modeling 9
However, they are aware that the experiment takes place in the context of a
thesis. For all subjects, anonymity is guaranteed. Before conducting the experi-
ment, for each level of social distance two pilot studies are performed to eliminate
ambiguities and misunderstandings as well as to improve modeling tasks. Fur-
ther, it is checked whether the social distance between the tasks is sufficiently
large. Finally, an evaluation sheet is created to assess the level of construal by
analyzing quality, granularity, and structure of resulting process models.
5.2 Experiment Execution
The experiment is executed in a computer lab at Ulm University. All in all,
95 students and staff members participate. Due to spatial constraints, up to 10
subjects conduct the experiment at the same time and several sessions within a
period of two weeks are offered. Each session lasts about 60 minutes and runs as
follows: The procedure of the experiment is explained and worksheets with task
descriptions are handed out. Thereby, subjects are randomly assigned to one of
the subject groups (cf. Sect. 4.4). Then, subjects start playing or watching the
warehouse scenario. Subsequently, they fill out an initial questionnaire capturing
their modeling experience. This information is used to test whether subjects are
familiar with process modeling. Then, subjects are asked to model the warehouse
scenario based on their own experience and in a way they think it is appropriate.
Finally, subjects provide their rating for perceived quality and may give feedback.
5.3 Data Validation
In total, data is collected from 95 subjects. One of them is excluded due to
invalidity of the process model obtained, i.e., the process model differs substan-
tially from the postulated task description. Hence, 94 subjects are considered for
data analysis, i.e., 84 students and 10 staff members (with 33 female subjects).
Further, the median concerning familiarity with BPMN is 3, i.e., above aver-
age. Regarding confidence with understanding BPMN process models, a median
value of 3 is obtained. Perceived competence in creating BPMN models has a
median value of 3. All values are based on a 5-point Likert scale. Prior to the ex-
periment, subjects analyzed 19 process models and created 7 in average.1Since
all values range above average and subjects are familiar with process modeling,
we conclude that subjects fit to the targeted profile.
5.4 Threats to Validation
Apparently, the experiment conducted faces the limitation that we did not in-
volve and compare professional process modelers and IT experts from industry,
but prospective ones (i.e. students). Although various investigations have shown
that students are proper substitutes for professionals in empirical studies (e.g.
[37, 38]) the results for professionals may differ.
1The full data set can be found in http://bit.ly/1VB2aS3
10 Michael Zimoch, Jens Kolb, and Manfred Reichert
6 Data Analysis & Interpretation
Sect. 6.1 presents descriptive statistics of the data gathered during the experi-
ment. Sect. 6.2 discusses whether a data set reduction is needed. Sect. 6.3 tests
the hypotheses. Finally, Sect. 6.4 discusses results.
6.1 Data Analysis and Descriptive Statistics
Figure 6 displays box plots (i.e., median, min, and max values as well as 1st and
3rd quartiles) of measurements for the pragmatic, semantic, perceived, and syn-
tactic quality. Further, the items of semantic and perceived quality are combined
into an aggregated variable [39], i.e., validity &completeness and agreement of
subjects. As a prerequisite, all response variables must show high reliability. For
this purpose, Cronbach’s αis calculated.2For semantic quality, a Cronbach with
α= 0.84 and for perceived quality a Cronbach with α= 0.77 results.
Low LowHigh HighLow HighLowHigh
0
2
1
3
4
Semantic Perceived SyntacticPragmatic
5
68 8
0
1
2
3
4
Fig. 6. Measurements for Quality
Low HighLow High LowHigh
0
0.50
0.25
0.75
1
Separable Cyclic Diameter
High
Sequential
Low 0
15
30
45
60
Fig. 7. Measurements for Structure
As shown in Figure 6, process models created by subjects with low social dis-
tance present a better level of understanding and contain less syntactical errors.
Regarding high social distance, in turn, process models seem to give a better
account of the domain. Moreover, perceived quality does not differ between the
subject groups. Further, Figure 7 presents calculated values for the process model
structure. There are only minimal differences in process model structure between
the process models. However, the diameter shows a clear difference depending
on the level of social distance. Process models whose subjects show a high social
distance contain notable longer paths (median of 23 for low, 30.5 for high).
Figure 8 shows results related to the granularity of process models, i.e., num-
ber of activities,gateways,nodes,edges,total process elements, and possible ex-
ecution paths. As a result, process model granularity is higher if subjects have
a high social distance. Especially, differences in the numbers of total process
elements are large. Note that low social distance results have a median of 60,
whereas high social distance leads to a median of 82 process elements.
6.2 Data Set Reduction
In general, the results of statistical analyses depend on the quality of the in-
put data, i.e., faulty data might lead to incorrect conclusions. Therefore, it is
2According to [39], α > 0.6 acceptable reliability; 0.7 < α < 0.9 good reliability
Social Distance in the Process of Process Modeling 11
LowLow Low HighLow High LowHigh
0
40
20
60
80
#Gateways #Nodes #Edges #Elements
High
#Activities
100
120
High
#Paths
Low High
177
Fig. 8. Measurements for Granularity
important to identify outliers and to evaluate whether these shall be excluded.
Note that the latter might be critical due to potential loss of information. In
the experiment, several outliers can be identified, but we decide to not remove
them since we consider them as correct, not being the result of wrong modeling.
Hence, removing them would bias results.
6.3 Hypothesis Testing
Response Variable p-value
Pragmatic Quality H1,1
Level of Understanding <0.01 (<0.05)
Semantic Quality H1,2
Validity & Completeness <0.01 (<0.05)
Perceived Quality H1,3
Agreement of Subjects 0.410 (>0.05)
Syntactic Quality H1,4
Number of Syn. Errors 0.046 (<0.05)
Level of Granularity H1,5
Number of Activities <0.01 (<0.05)
Number of Gateways 0.039 (<0.05)
Number of Nodes <0.01 (<0.05)
Number of Edges <0.01 (<0.05)
Number of Elements <0.01 (<0.05)
Number of Paths 0.148 (>0.05)
Process Model Structure H1,6
Sequentiality 0.326 (>0.05)
Separability 0.092 (>0.05)
Cyclicity 0.258 (>0.05)
Diameter <0.01 (<0.05)
Table 1. Results of Hypotheses Testing
Sect. 6.1 indicates differences regard-
ing low and high social distance. In the
following, we test whether observed
differences are statistically significant.
We test the response variables with the
Mann-Whitney-U-test [40]. A success-
ful u-test (with p < p0at risk level
α= 0,05) will reject a null hypoth-
esis. Table 1 shows the results of hy-
pothesis testing (cf. Sect. 4.2). In sum-
mary, hypotheses H1,1,H1,2, and H1,4
can be accepted. Despite the number
of significant results, like H1,6,H1,5
is only partially supported, and thus
both hypotheses cannot be accepted.
In addition, H1,3shows no significance
and, hence, must be rejected. Based on
the results, we may conclude that so-
cial distance (i.e., event which is not
self-experienced) leads to a change in
the quality,granularity, and structure
of resulting process models.
6.4 Discussion
The results indicate that process designers showing a high social distance (i.e.
passive participation) to a particular business process tend to create a more fine-
12 Michael Zimoch, Jens Kolb, and Manfred Reichert
grained, detailed, and complete process model, i.e., reflecting a high semantic
quality and granularity. In turn, process designers showing a low social distance
(i.e. active participation) create a more course-grained and abstract, but easy
to understand process model with less syntactical errors, i.e., reflecting a high
pragmatic and syntactic quality. Regarding perceived quality and process model
structure, final results do not show any or only small differences.
Interestingly, the results only partially comply with CLT (cf. Sect. 2) and
our previous experiment [19]. It appears that the investigated factor of the social
distance (cf. Sect. 4) has a different impact on the process of process modeling
and, hence, resulting outcomes differ in several aspects (cf. Sect 6.1). As possible
explanation an active participation results in major attention devoted to actions
performed by oneself, while a passive participation results to equal attention paid
to all details [41]. BPMN knowledge might be a critical moderator reversing the
relationship between construal level and distance (i.e., social distance) leading
to circumstances where the abstract seems near and the concrete seems far [42].
However, combining previous results, in general, one can assume that social
distance leads to a change in the quality, granularity, and structure of resulting
process models. It is noteworthy that results differ depending on how a pro-
cess designer experiences social distance, i.e., relation to other individuals or
events which are not self-experienced. For enterprises, it is thus recommended
to evaluate the modeling domain and, hence, to involve specific process design-
ers to ensure desired outcomes; e.g., to achieve a high process model quality,
it is thus recommended to involve process designers being more confident with
corresponding business processes.
7 Related Work
This paper investigates the impact of social distance on the quality, granular-
ity, and structure of process models. The work is related to frameworks and
guidelines dealing with process model quality. SEQUAL uses semiotic theory for
identifying various aspects of process model quality [25], whereas GoM describes
quality considerations for process models [7]; 7PMG, in turn, characterizes desir-
able properties of a process model [8]. Moreover, research on comprehensibility
and maintainability exists. The influence of model complexity on process model
comprehensibility is investigated in [5]. [35] discusses factors for errors in process
models; [43] discusses the impact of different quality metrics on error probability.
[44] provides prediction models for true usability and maintainability of pro-
cess models. How and at which level of granularity a designer models a particular
process is described in [26]. In the context of process modeling only little work
exists that takes cognitive aspects into account. [28] presents the effects of re-
ducing cognitive load on end user understanding of conceptual models, whereas
[11] describes the cognitive difficulty of understanding different relations between
process model elements.
Common to all these approaches is the focus on the created process model
(i.e., the product of process modeling), while little attention has been paid on
the process of the process modeling itself. Nautilus complements related work by
Social Distance in the Process of Process Modeling 13
investigating the process of process modeling for tracing model quality back to
modeling strategies resulting in process models of different quality [45].
The effectiveness of gamification based on a quality service model analyzing
the social and psychological motivations of participants is discussed in [46]. Agile
and efficient responds to changing requirements and consequential amendments
to corresponding business processes are provided in [47], based on a gamification
and BPM approach incorporated into a social network. Finally, [48] provides
preliminary evidence that blending process management to gamification concepts
may be beneficial.
Considerable work involving conceptual modeling of processes in a 3D virtual
world can be found in [49]. In addition, [50] provides an approach for collabo-
rative process modeling using a 3D environment. A similar use case in a 3D
scenario to visualize storyboards for business process models is proposed in [51].
8 Conclusion
This paper investigated whether social distance affects the process of process
modeling and its outcomes, i.e., the quality, granularity, and resulting process
model structure. In particular, an experiment using gamification in a virtual
world was conducted showing that there are significant differences depending
on whether a process designer has a low or high social distance to the modeled
domain. While first results look promising, further investigations are desirable.
More precisely, their generalization needs to be confirmed by additional empirical
experiments to obtain more accurate results allowing for such a generalization.
As a next step, we will focus on psychological distance (i.e., social, spatial,
temporal, hypothetical) as well as the use of gamification and virtual worlds
to learn more about the particular effects on the process of process modeling.
Combining experiment results enables us to extract guidelines on how modeling
teams in enterprises should be composed and optimal process models can be
obtained. Finally, experiments with practitioners are planned to validate results
in real-world scenarios.
References
1. Weber, B., Reichert, M., Mendling, J., Reijers, H.A.: Refactoring Large Process
Model Repositories. In: Computers in Industry. (2011) 467–486
2. Mendling, J.: Metrics for Process Models: Empirical Foundations of Verifiation,
Error Prediction, and Guidelines for Correctness. Springer (2008)
3. Reijers, H.A, Mendling, J: A Study into the Factors that Influence the Under-
standability of Business Process Models. In: Systems, Man and Cybernetics, Part
A: Systems and Humans. (2011) 1–14
4. Moody, D.L.: Theoretical and Practical Issues in Evaluating the Quality of Con-
ceptual Models: Current State and Future Directions. In: Data & Knowledge
Engineering. (2005) 243–276
5. Mendling, J., Reijers, H.A., Cardoso, J.: What Makes Process Models Understand-
able? In: Proc 5th Int’l Conf on Business Process Mgmt. (2007) 48–63
6. Mendling, J., Strembeck, M.: Influence Factors of Understanding Business Process
Models. In: Proc 11th Int’l Conf on Business Informations Systems. (2008) 142–153
14 Michael Zimoch, Jens Kolb, and Manfred Reichert
7. Becker, J., Rosemann, M., Uthmann, C.V.: Guidelines of Business Process Mod-
eling. In: BPM, Models, Techniques, and Empirical Studies. (2000) 30–49
8. Mendling, J., Reijers, H.A., van der Aalst, W.M.: Seven Process Modeling Guide-
lines (7PMG). In: Inf & Software Techn. (2010) 127–136
9. BPM Research Cluster: Cheetah Experimental Platform (2016) http://bpm.q-
e.at/?page id=56.
10. Figl, K., Weber, B.: Individual Creativity in Designing Business Processes. In:
Advanced Information Systems Engineering Workshops. (2012) 294–306
11. Figl, K., Laue, R.: Cognitive Complexity in Business Process Modeling. In: Ad-
vanced Information Systems Engineering. (2011) 452–466
12. Trope, Y., Liberman, N., Wakslak, C.: Construal Levels and Psychological Dis-
tance: Effects on Representation, Prediction, Evaluation, and Behavior. In: Journal
of Consumer Psychology. Volume 17. (2007) 83–95
13. Todorov, A., Goren, A., Trope, Y.: Probability as a Psychological Distance: Con-
strual and Preferences. In: J of Exp Social Psychology. Volume 43. (2007) 473–482
14. Trope, Y., Liberman, N.: Construal-level Theory of Psychological Distance. In:
Psychological Review. Volume 117. (2010) 440–463
15. Day, S., Bartels, D.: Representation over Time: The Effects of Temporal Distance
on Similarity. In: Cognition. Volume 106. (2008) 1504–1513
16. Liberman, N., Sagristano, M.D., Trope, Y.: The Effect of Temporal Distance on
Level of Mental Construal. In: J of Exp Social Psych. Volume 38. (2002) 523–534
17. Fujita, K., Henderson, M.D., Eng, J., Trope, Y., Liberman, N.: Spatial Distance
and Mental Construal of Social Events. In: Psych Sci. Volume 17. (2006) 278–82
18. Pronin, E., Olivola, C.Y., Kennedy, K.A.: Doing Unto Future Selves As You Would
Do Unto Others: Psychological Distance and Decision Making. In: Personality &
Social Psychology Bulletin. Volume 34. (2008) 224–36
19. Kolb, J., Zimoch, M., Weber, B., Reichert, M.: How Social Distance of Process
Designers Affects the Process of Process Modeling: Insights from a Controlled
Experiment. In: Proc SAC’14. (2014) 1364–1370
20. Deterding, S., Dixon, D., Khaled, R., Nacke, L.: From Game Design Elements to
Gamefulness: Defining ”Gamification”. In: MindTrek’15. (2011) 9–15
21. Zichermann, G., Cunningham, C.: Gamification by Design: Implementing Game
Mechanics in Web and Mobile Apps. O’Reilly (2011)
22. Bartle, R, A.: Designing Virtual Worlds. New Riders (2004)
23. Davis, A., Khazanchi, D., Murphy, J., Zigurs, I., Owens, D.: Avatars, People, and
Virtual Worlds: Foundations for Research in Metaverses. In: J of the Association
for Information Sys. Volume 10. (2009) 90–117
24. Moody, D.L.: The ”Physics” of Notations: Toward a Scientific Basis for Construct-
ing Visual Notations in Software Engineering. In: Soft Eng. (2008) 756–779
25. Krogstie, J.: Model-Based Development and Evolution of Information Systems.
Springer (2012)
26. Holschke, O., Rake, J., Levina, O.: Granularity as a Cognitive Factor in the Effec-
tiveness of Business Process Model Reuse. In: LNCS. (2009) 245–260
27. Polyvyanyy, A., Smirnov, S., Weske, M.: On Application of Structural Decompo-
sition for Process Model Abstraction. In: Proc BPSC’09. (2009) 110–122
28. Moody, D.L.: Cognitive Load Effects on End User Understanding of Conceptual
Models: An Experimental Analysis. In: Advances in Databases and Information
Systems. (2004) 129–143
29. Wolf, C., Harmon, P.: The State of Business Process Management 2012. In:
BPTrends Report. (2012)
Social Distance in the Process of Process Modeling 15
30. Petre, M.: Why Looking Isn’t Always Seeing: Readership Skills and Graphical
Programming. In: Commun ACM. Volume 38. (1995) 33–44
31. Lindland, O.I., Sindre, G., Solvberg, A.: Understanding Quality in Conceptual
Modeling. In: IEEE Software. (1994) 42–49
32. Krogstie, J., Sindre, G., Jørgensen, H.: Process Models Representing Knowledge
for Action: A Revised Quality Framework. In: J Inf Syst. Volume 15. (2006) 91–102
33. Recker, J., Safrudin, N., Rosemann, M.: How Novices Model Business Processes.
In: Proc 9th Int’l Conf on Business Process Management. (2010) 29–44
34. Rittgen, P.: Quality and Perceived Usefulness of Process Models. In: Proc 24th
Symp on Applied Computing (SAC’10). (2010) 65–72
35. Mendling, J., Neumann, G.: Error Metrics for Business Process Models. In: Proc
CAISE’07. (2007) 53–56
36. Wohlin, C., Runeson, P., H¨ost, M., Ohlsson, M.C., Regnell, B., Wesslen, A.: Ex-
perimentation in Software Engineering - An Introduction. Kluwer (2000)
37. H¨ost, M., Regnell, B., Wohlin, C.: Using students as subjects a comparative study
of students and professionals in lead-time impact assessment. Empirical Software
Engineering 5(2000) 201–214
38. Svahnberg, M., Aurum, A., Wohlin, C.: Using students as subjects - an empirical
evaluation. In: ESEM’08, ACM (2008) 288–290
39. Kline, P.: Handbook of Psychological Testing. Volume 2. Routledge (1999)
40. Sirkin, M.: Statistics for the Social Sciences. Volume 3. Sage (2005)
41. Tversky, A., Kahnemann, D.: Availability: A Heuristic for Judging Frequency and
Probability. In: Cognitive Psychology. Volume 5. (1973) 207–232
42. Kyung, E.J., Menon, G., Trope, Y.: Construal Level and Temporal Judgments of
the Past: the Moderating Role of Knowledge. In: Psychonomic Bulletin & Review.
Volume 21. (2014) 734–739
43. Mendling, J., Verbeek, H.M.W., van Dongen, B.F., van der Aalst, W.M.P., Neu-
mann, G.: Detection and Prediction of Errors in EPCs of the SAP Reference
Model. In: Data and Knowledge Eng. (2008) 312–329
44. Aguilar, E.R., Sanchez, L., Carballeira, F.G., Ruiz, F., Piattini, M., Caivano, D.,
Visaggio, G.: Prediction Models for BPMN Usability and Maintainability. In: Proc
CEC’09. (2009) 383–390
45. Pinggera, J., Soffer, P., Fahland, D., Weidlich, M., Zugal, S., Weber, B., Reijers, H.,
Mendling, J.: Styles in Business Process Modeling: an Exploration and a Model.
In: Soft & Sys Modeling. (2013) 1–26
46. Aparicio, A.F., Vela, F.L.G., S´anchez, J.L.G., Montes, J.L.I.: Analysis and Applica-
tion of Gamification. In: New Trends in Interaction, Virtual Reality and Modeling,
Human-Computer Interaction Series. (2013) 113–126
47. Santorum, Marco and Front, Agnes and Rieu, Dominique: ISEAsy: A Social Busi-
ness Process Management Platform. In: Proc BPM’13. (2013) 125–137
48. Brito, Thaigo P. and Paes, Josias and Moura, Antao B.: Game-Based Learning in
IT Service Transition. In: Proc CSEDU’14. (2014) 110–116
49. Brown, Ross A.: Conceptual Modelling in 3D Virtual Worlds for Process Commu-
nication. In: Proc APCCM’10. (2010) 25–32
50. West, Stephen and Brown, Ross A. and Recker, Jan C.: Collaborative Business
Process Modeling Using 3D Virtual Environments. In: Proc APCCM. (2010) 51–60
51. Kathleen, Nardella and Brown, Ross A. and Kriglstein, Simone: Storyboard Aug-
mentation of Process Model Grammars for Stakeholder Communication. In: Proc
IVAPP’14. (2014) 114–121
