Aligning stakeholders’ mental models on carsharing system using remote focus group method

Abstract

The novelties of new mobility solutions, such as carsharing, may instill different expectations and understanding of the concepts among stakeholders. These differences in their ‘mental models’ can hamper the wider implementation of the concept and delay a transition toward a more sustainable transport system. In this study, we implemented a participatory group modeling building approach (GMB) to explore the differences and to integrate the mental models of stakeholders concerning the carsharing operation in Bangkok, Thailand. Through the process, we identified apparent differences in how participants visioned a successful carsharing operation and created an initial shared understanding in the form of a causal loop diagram. The qualitative model included attributes influencing the success of carsharing and possible policy interventions. The results illustrated the effectiveness of GMB as a participatory approach for transport planning.

Full text

Transportation Research Part D 101 (2021) 103122
Available online 23 November 2021
1361-9209/© 2021 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license
(http://creativecommons.org/licenses/by/4.0/).
Aligning stakeholders’ mental models on carsharing system using
remote focus group method
Peraphan Jittrapirom
a
,
b
, Saroch Boonsiripant
c
,
*
, Monthira Phamornmongkhonchai
c
a
Nijmegen School of Management, Radboud University, Nijmegen, the Netherlands
b
Center for Global Environmental Research, National Institute for Environmental Studies, Tsukuba, Japan
c
Department of Civil Engineering, Kasetsart University, Bangkok 10900, Thailand
ARTICLE INFO
Keywords:
Smart mobility
Group model building
Participatory approach
Transport planning
ABSTRACT
The novelties of new mobility solutions, such as carsharing, may instill different expectations and
understanding of the concepts among stakeholders. These differences in their ‘mental models’ can
hamper the wider implementation of the concept and delay a transition toward a more sustain-
able transport system. In this study, we implemented a participatory group modeling building
approach (GMB) to explore the differences and to integrate the mental models of stakeholders
concerning the carsharing operation in Bangkok, Thailand. Through the process, we identied
apparent differences in how participants visioned a successful carsharing operation and created
an initial shared understanding in the form of a causal loop diagram. The qualitative model
included attributes inuencing the success of carsharing and possible policy interventions. The
results illustrated the effectiveness of GMB as a participatory approach for transport planning.
1. Introduction
Although cities and urban regions around the world are committing themselves to climate targets and sustainable development
goals, the transition toward a more sustainable and low-carbon future remains slow. In particular, in the transport sector, emissions
continue to increase and numerous urban areas are still faced with the negative effects of transport, such as severe congestion, air
pollution, and transport accidents (European Commission, 2019). In order to stimulate a transition toward a sustainable transport
system, a consistent and integrated commitment is required at all levels of governance involving concerted efforts from all stakeholders
in the system, such as decision-makers, service providers, and travelers (Stephenson et al., 2018).
Emerging mobility services and solutions, such as carsharing and Mobility-as-a-Service (MaaS), have the potential to address urban
transport challenges but their adoption and implementation have been limited, particularly in developing countries where the
importance and urgency regarding the sustainability of the transport system are relatively high (Jackson et al., 2019; Lane et al., 2015).
There are reasons behind the stagnation. In particular, different stakeholders in the transport system may have different levels of
understanding and expectation (mental models) concerning the novel mobility concepts (the concepts may be poorly understood and
have unclear implications) that hinder their implementation (Beutel et al., 2014; Jittrapirom et al., 2018a). Other research in car-
sharing and MaaS has revealed the complexity of their operations, primarily due to the multiple entities and the stakeholders con-
nected to and involved in them (Jittrapirom et al., 2017; Jorge and Correia, 2013).
A mental model is dened as an internal conceptual representation of an external system (Doyle and Ford, 1988). It is a set of causal
* Corresponding author.
E-mail address: saroch.b@ku.th (S. Boonsiripant).
Contents lists available at ScienceDirect
Transportation Research Part D
journal homepage: www.elsevier.com/locate/trd
https://doi.org/10.1016/j.trd.2021.103122

Transportation Research Part D 101 (2021) 103122
2
relationships that each individual holds on a certain matter, which is based on knowledge and previous experience. These mental
models are a useful simplication to support decision making but are also imperfect like other types of model (Sterman, 2000). The
concept is rooted in cognitive psychology and has been applied to explain the thinking, learning, and decision-making processes of
individuals and groups. Research in this eld has shown that the alignment of mental models of individuals within a group can in-
uence the quality of the group’s decision-making and its performance (Lim and Klein, 2006).
In the eld of transport research, the concept of mental models is primarily applied to understand types of traveler and driver
behavior (Pampel et al., 2015). However, involving stakeholders in the policy-making process can be seen as a way to elicit and
incorporate elements of stakeholders’ mental models into the decision-making process. Stakeholder engagement in transport planning
and the decision-making process has several recognized benets as it can widen perspectives into the subject, enhance the insights and
information included, and induce acceptance and commitment toward the outcomes reached (Leyden et al., 2017; Marcucci et al.,
2018; May, 2015).
Past studies have explored how stakeholders can be engaged using different approaches, such as the stated-preference gamication
(Marcucci et al., 2018) and the analytic hierarchy process with agent-based modeling (Le Pira et al., 2018), to elicit stakeholder
preference and to stimulate reaching consensus in the transport planning process. However, these approaches tend to be quantitative in
nature as well as resource/data intensive. Furthermore, the cases applied were types of transport projects with which the stakeholders
tended to be more familiar (such as mobility management and cycling promotion strategy), particularly as they were set in the context
of developed countries. In contrast, in developing countries, the data required by these approaches may not be available, particularly
for a planning process concerning new mobility concepts, such as carsharing and MaaS, where past information or relevant historical
data are likely to be scarce.
The objectives of this research were threefold. First, it aimed to examine how relevant stakeholders have different underlying
assumptions and hold different viewpoints and expectations concerning the implementation of the carsharing concept. Second, it
examined how the qualitative participatory modeling approach in group model building (GMB) can integrate these perspectives. GMB
has been applied to provide insights into complex systems and has been an effective method to support stakeholder engagement (see
van Bruggen et al., 2019). Third, using the participatory approach in the GMB technique during the outbreaks of coronavirus, we
explored how GMB, which is typically implemented in a face-to-face workshop format, can be carried out remotely to minimize the risk
of exposure to the coronavirus pandemic for the participants and the research team.
Our study looked to complement other studies that used a quantitative approach to support stakeholder engagement in transport
planning and also carsharing studies that typically examined the concept from a single perspective, such as a user’s preference or
carsharing operation (Schreier et al., 2018; Shaheen and Cohen, 2007). The process was applied here to examine the system
perspective of a carsharing operation by bringing together the relevant stakeholders, such as policymakers, regulators, operators,
researchers, user representatives, and insurance providers, to exchange their knowledge and expertise. We expected the study to
produce a qualitative system structure of the city’s carsharing operation for policy evaluation and to identify planning elements (such
as vision, key performance indicators, and barriers to implementation) for policy formulation. We expected the participants to gain a
new understanding and insights into the urban carsharing system. Additionally, the application of this ‘bottom-up’ approach to support
the implementation of shared mobility concepts can be instrumental, particularly, in the domain of transportation planning in which
public institutions take dominant roles in formulating and implementing plans.
It should be noted that our study was exploratory in nature, seeking to investigate the different mental maps among the stake-
holders and to show what a comprehensive shared image of a complex problem (in our case; implementing carsharing for Bangkok)
would look like if we combined the worldviews of various types of stakeholders. The approach is comparable to focus group research
with each stakeholder committing 7–8 h of their time in total.
The remainder of this paper is organized as follows. Section 2 describes the carsharing background and its implementation in
Bangkok. In Section 3, we present the research method, on how remote GMB was prepared and carried out. In Section 4, we present and
discuss the results from the exercise and reect on the process and its limitations. Lastly, we conclude the study in Section 5 with a
summary of the key takeaway messages and highlight opportunities for future research.
2. Carsharing implementation in Bangkok city, Thailand
By distancing vehicle usage from ownership, urban carsharing enables individuals to gain the benets associated with access to a
car without the nancial burden and responsibilities related to the ownership (Shaheen et al., 2019). This access-based mobility has
been linked with benets associated with urban sustainability, such as the reduction and delay of private vehicle ownership and
vehicle miles traveled, promoting public transport usage, enhancing accessibility, and reducing energy consumption and greenhouse
gas emissions (Giesel and Nobis, 2016; Kent and Dowling, 2013; Schreier et al., 2018; Shaheen et al., 2019). Implementation and
adoption of carsharing have been gaining momentum in recent years but primarily in the developed countries within North America
and Western Europe, such as the USA, Canada, Italy, and Germany (Shaheen and Cohen, 2013). In contrast, there has been limited
adoption and implementation in developing countries, such as China and India (Movmi, 2019).
Past efforts to promote carsharing in developing countries have met with obstacles, such as a lack of public awareness of the
service’s presence, social aspiration to personal car ownership as a sign of status, and severe congestion that leads to unpredictable
travel times (and thus rental cost) for users (Shaheen and Martin, 2010). Other hindrances include the existence of more affordable
point-to-point transport services (including rickshaw, taxi, and ride-hailing services) and the lack of support from the public sector in
the form of policy and legislation, such as allocated parking spaces for carsharing to reduce operation costs (Lane et al., 2015; Movmi,
2019). Several of these challenges also stem from the car-sharing concept being unfamiliar to public authorities and regulators in these
P. Jittrapirom et al.

Transportation Research Part D 101 (2021) 103122
3
countries, as it is still an emerging and novel option for many local authorities (Lane et al., 2015; Münzel et al., 2018). Additionally, the
complexity of a carsharing operation which involves multiple entities and players, all of whom are interconnected, can also hinder
stakeholder understanding of the concept (Jorge and Correia, 2013; Spickermann et al., 2014).
Similar to other developing countries, the carsharing concept is in its infancy in Thailand. Its rst carsharing service started
operation in Bangkok, the capital city, in 2014. Presently, there are an estimated 4,000 carsharing vehicles in the country, operated by
four providers, namely Haupcar (350 vehicles, station-based service), Toyota Ha:mo (30 vehicles, one-way service), Drivemate (3,500
vehicles, Peer-to-Peer service), and ASAP Go (100 vehicles, Business-to-Business service) (ASAP, 2018; CU, 2018; Manager Online,
2020).
Despite the emergence of carsharing in Thailand, regulations, and laws are still lagging and ambiguous about the position of these
services. For example, current Thai laws do not require car rental and carsharing companies to obtain any business license or permit
from the government. As a result, individuals and local businesses such as guest houses, local tour guides, and convenience stores, also
provide car rental services without any regulations from the government. (Park and Suttijaree, 2017). The unregulated providers have
also caused several other issues to arise such as unfair vehicle rental agreements, uninsured rental vehicles, and fraudulent cases.
In recent years, some efforts have been made by different organizations to address the conditions mentioned. For example, in 2014,
the Ministry of Tourism and Sports along with the Thai Car Rental Association (TCRA) proposed a Motor Vehicle Rental Bill to regulate
the car rental industry and to protect both the car rental companies and travelers. The Ofce of Insurance Commission (OIC) also began
to promote innovative insurance products by implementing a regulatory sandbox that led to the development of pay-per-use insurance
plans. Such a exible insurance service is highly attractive for Peer-to-Peer shared vehicle owners. Additionally, the Ofce of Transport
and Trafc Policy and Planning (OTP) and the Bangkok Metropolitan Administration have recognized carsharing as a solution to
transport challenges in Bangkok city (MOT, 2017). This acknowledgment opened possibilities for supporting policies and regulations.
Several private Thai residential developers have acquired a eet of electric vehicles and collaborated with carsharing operators to
provide mobility solutions for their customers and tenants (Bangkok Post, 2018). However, these measures are often fragmented and
may stem from how each organization perceives the challenges from their point of view without any system perspective. These
piecemeal efforts are unlikely to address the critical challenges Bangkok carsharing is facing (such as low public awareness, parking
availability, and unfair price competition with other modes of transport) and can also have unintended effects on the urban system.
3. Research method - remote group model building
GMB refers to a process to engage stakeholders in the development of a system dynamics model that supports understanding
complex systems (Rouwette and Vennix, 2019). Through a series of facilitated and scripted activities, the stakeholders involved can
exchange their perceptions of facing problems and of the possible underlying causes (Andersen et al., 1997b). Broadly, the goals of
GMB are to support individuals in their social learning through the renement of their mental models. Collectively at the group and
organizational levels, it aims to align the shared mental models of the group to enhance consensus and commitment to a decision that
results in changed behavior and system changes in the organization (Andersen et al., 1997b; Rouwette et al., 2011). The use of the
model is primarily as a means to communicate and to integrate various ideas about a problem (Quade, 1972). Through involving the
stakeholders of a problem in this co-creation process, the model also becomes a learning tool that supports the building of shared vision
and understanding of the subject (de Geus, 1988; Lane, 1992; Senge, 1990). Other studies have highlighted the benet of GMB as an
effective method to enhance insight into a problem, to improve group communication, and to support a shared vision and strategy
formation (Rouwette et al., 2011; Vennix, 1996).
The system approach toward problem structuring in GMB and its aims make the approach akin to other facilitated approaches such
as soft system methodology (Mingers and Sarah, 1992) and focus groups (Kitzinger, 1995). However, GMB leverages the diagramming
convention of causal loop diagramming (CLD), making it an effective method to capture, elicit, and communicate the mental models of
individuals and collectives. In addition, GMB is versatile as it can take shape in different variations depending on factors such as
expected results (a qualitative or quantitative model) or a required format (structured or unstructured) (Hovmand, 2014; Vennix,
1996).
Given these prospects for the GMB approach, there are several challenges concerning its applications and in achieving the goals. For
example, there is a large variation in the facilitation ‘techniques’ being applied to cases that makes it difcult to compare studies or to
evaluate the GMB effects objectively due to the lack of studies that have evaluated their outcomes against control cases (Rouwette
et al., 2002). Additionally, an application of the GMB process can be resource-intensive. Some of these shortfalls may be attributed to
factors that are beyond the control of the GMB researchers, such as the difculty of fully understanding a client’s problem in advance of
the intervention, the unknown group dynamics among stakeholders, and the difculty in identifying what has caused any learning and
change behavior (see for example Andersen et al., 1997b). Several researchers have sought to improve the technique by strengthening
the objectiveness of the method by formulating scripts (Andersen et al., 1997a; Andersen and Richardson, 1997) and enhancing the
scripts to improve GMB’s effectiveness in problem structuring (Ackermann et al., 2010).
GMB has been applied to a wide range of complex and contested issues, such as the food, water, energy nexus, the construction of
resource, and environmental models (Purwanto et al., 2019; Voinov et al., 2016), and to support the dialogue process on sustainability
(Ruth et al., 2015). In addition, its application to transport and smart community planning has been explored (Jittrapirom et al., 2017;
Rees et al., 2017; Yoshida et al., 2020 and in reviews such as Rouwette et al., 2002 and Scott et al., 2016). However, applications are
still limited that support the implementation of novel mobility concepts such as carsharing. One related study by Esfandabadi et al.
(2020) took a systematic approach and utilized CLD to construct a conceptual framework on the interconnections between carsharing
services and their environmental effects; however, the study was based on a literature review.
P. Jittrapirom et al.

Transportation Research Part D 101 (2021) 103122
4
Typically, GMB is carried out in a face-to-face setting by one or more facilitators who may assume different roles at different points
in the workshop (Luna-Reyes et al., 2006). In moving the GMB process online, the methodological approach of GMB in the current
study involved four steps: 1) knowledge elicitation, 2) conceptualization of the system and the interconnectedness among variables, 3)
identication and assessment of possible interventions, and 4) evaluation of the process (see Akkermans and Rouwette (1996) for the
detailed GMB steps). The knowledge elicitation was carried out through semi-structured interviews with stakeholders (Galletta and
Cross, 2013), which took place during May–July 2020. We selected the semi-structured interview because its exible format allowed
us to discuss the research process, understand the participants’ background and existing perception of carsharing, and build personal
connections with the respondents.
For the convergent activities (steps 2 and 3), a collaborative platform allows synchronized viewing of a causal map and supports
interactive discussion. We selected Zoom and Miro and modied the divergent standard scripts from Scriptapedia (2020), an open-
Table 1
Summary of interview transcripts classied by stakeholders’ sector.
Sector What does a successful carsharing
operation look like? (with suggested KPIs)
What factors contribute to the success if
present (+) and hinder success if lack (-)?
What are known existing policies and
proposed policies that affect carsharing
operation?
1.Policymakers and
public sectors
Aligns with the national Smart Mobility
plan; Stimulates GDP growth; Increases
employment; Reduces costs, emission, and
energy consumption; Provides a safe,
convenient transport system; Enhances
accessibility to transportation service,
Protects personal data privacy
Suggested KPIs: No. of users; Car ownership
per capita
(þ): Infrastructures provisions (e.g.,
parking and special lane); Good integration
with the public transport system; Aversion
to mass transit due to COVID-19; Easy
access, affordable and convenient to use;
High quality of the service;
(-): Highly attractive compared to other
modes; Affordable costs compared to other
available modes;
(+/-): Social awareness and acceptance
Existing: Public transport promotion;
Reduced tax on innovation investment;
Pilot on-demand driving insurance
Proposed: Shared mobility and tourism
promotion policies; Investment in big data
management and e-payment; Innovation
promotion
2.Representatives of
users
Convenience and exibility with a variety
of services, e.g., one-way service; Supports
P2P rental service; High number of users;
Limited need for private cars
Suggested KPIs: No. of users, private car
sales; Fleet size; Public awareness of the
service
(þ): Several providers in the market; Lower
private car ownership;
(-): Affordable price of the service;
Sufcient number of vehicles; Condence
in cleanliness and quality;
(þ/-): High public awareness; Supportive
measures and regulations from the
government;
Existing: None
Proposed: Supportive legislation and
regulations e.g., tax, collaboration among
private sector; Promotion of competitions
and avoid monopoly; Marketing activity to
raise public awareness
3.Researcher and
smart mobility
experts
A viable alternative mode of transportation
that offers convenience and privacy to its
users; Easy to access, reserve, ubiquitous,
and has good coverage
Suggested KPIs: Level of accessibility,
convenience, quality, functionality, and
availability
(þ): Supporting parking policy; Integration
with other transport services and
businesses; Service characters (e.g., one-
way) that provide convenience to user
(-): Meeting the needs of users; Reasonable
number of potential users; Affordability,
convenience, and ease of use
(þ/-): Public awareness; Quality of public
transport
Existing: Public transport promotion
Proposed: Provision of exclusive parking
and fare exemption; Strong cooperation
with private sector players.
4.Carsharing
providers
A protable and sustainable business;
Meets the needs of users; Popular; An
integral part of daily life; High variety of
available features, e.g., one-way, EV, e-
scooter; Extended coverage.
Suggested KPIs: Utilization rate; No. of users
and organizations that utilize services;
Changes in user behavior, e.g., reduction in
private car ownership
(þ): Emerging technology that lowers
operational costs
(-): Responsible consumer behavior;
Reasonable operational costs; Good
integration with public transport service
(þ/-): Governmental support; Enabling
conditions, e.g., available parking space
and lower congestion; Collaboration to
share resources, exchange knowledge, and
create new business models; Public
awareness and understanding; a shift in
social value on car ownership
Existing: None
Proposed: Supporting policies on parking,
tax, and toll exemptions; Supporting new
technology (e.g., electric vehicles); Public
promotions; Integration with public
transport as a rst-and-last-mile solution
and other business services; Restriction of
personal car ownership and car use
5.Private
organizations
Self-sustained industry; Meets the needs of
users and provides a ubiquitous, high
quality, and safe transport service; Make
positive societal contributions; Raise public
awareness and understanding of the
sharing economy.
Suggested KPIs: No. of users; Service
revenue; Level of availability; Car
ownership
(þ): Increase awareness on environmental
sustainability; Shifting of social values on
car ownership among young people;
Understanding of carsharing services;
Supporting infrastructure
(-): Familiarity of users; Condence in
vehicle safety; Policy, social value, and
habit that promote private car ownership
(þ/-): Affordability; Supportive public
policies and an ecosystem; Cooperation
between sectors
Existing: Tax incentive for personal car
purchase and high electric vehicle import
tax
Proposed: Initiative to support sharing
economy; Supportive regulations, subsidy,
and tax deductions to support operation
and promote competitions in the market;
Parking regulation; Public transport
development and integration with
carsharing; Cooperation to implement
carsharing; Effective management of
operation; Inclusion of electric vehicles in
the eet
Note: KPIs – Key Performance Indicators.
P. Jittrapirom et al.

Transportation Research Part D 101 (2021) 103122
5
source GMB script for structured exercise, to accommodate the constraints imposed by the remote environment. The script selection
was based on the experience of the facilitators and the purposes and intended outcomes of the exercise. The GMB sessions were kept
concise to minimize any concentration fatigue by the respondents. These activities took place during a two-session online workshop
(August 5 and 26, 2020), and the nal step took place through an online follow-up survey (during August 26–September 2, 2020). To
our knowledge, only a recent study by Wilkerson et al. (2020) carried out a GMB exercise in a similar setting. Our reections on
adapting GMB activities into an online format are included in Section 4.4.
At the outset, we identied the stakeholders through discussion with a group of experts who had experience regarding the car-
sharing service industry in a city. Care was taken to identify the apparent inuences and connections of these stakeholders to the
system (Vennix, 1996). These stakeholders are also typically invited by the government to take part in public consultation concerning
transport policy as some of them are directly involved with the governance and operation of the city’s transport system (such as the
insurance regulator). We also asked each interviewee to recommend additional contacts (“snowballing”). The nal respondent list
comprised 23 stakeholders consisting of the public sector and governmental agencies (11), service providers (4), the private sector (3),
smart mobility community and research (3), and users (2). All stakeholders agreed to participate in the interviews though four par-
ticipants from a public organization refrained from joining the workshops. The rst workshop was attended by 15 respondents and the
second workshop by 17 respondents (see Appendix A for details). The names and organizations of these participants were excluded to
protect the anonymity of the participants.
For the semi-structured interviews, we sent the respondents the questionnaire in advance. It asked them to describe: 1) their roles
within the Bangkok transport and carsharing systems, 2) their vision of a successful carsharing system, 3) factors that might hinder or
accelerate the process toward the implementation of the vision, and 4) policies or measures that should be considered. All interviews,
except three, were conducted and recorded via an online teleconference service. The other three interviews were carried out in person
at each interviewee’s request and were conducted in an open space or a large room, with all participants wearing facemasks during the
interviews. On average, each interview lasted 30 min and for the organizations that had more than one participant, the interviews were
carried out in group settings. The interviews were then transcribed and translated into English and used to construct a CLD for each of
them to illustrate the points discussed. The interviewees were sent their transcripts and personal CLDs to conrm their accuracy. The
transcripts were then open-coded (Strauss and Corbin, 1990), classied, and analyzed to provide a list of entities to be considered in
the workshop sessions (see Table 1).
Before the workshops, the research team prepared and made available a shared space on Miro for the participants. The shared space
was divided into three sections: Pre-workshop, Workshop 1, and Workshop 2. It included information related to the workshop, such as
the program, the presentations, personal CLDs and other outputs from the interviews, links to relevant information, and shared
working space for the workshop days. The participants were also expected to carry out simple tasks (ll in the self-introduction form
and share their expectations) to familiarize themselves with Miro. During the sessions, the facilitators would direct the participants to
different parts of the board.
In the rst workshop (3 h), after a brief welcome and an explanation of the goal of the exercise, the research team (one main
facilitator and four group facilitators) and the participants introduced themselves. The team explained the workshop process and the
iconography or the signs and symbols of CLD (Richardson, 2013) and split the participants into four groups. Each group had 4–5
members from different sectors and a group facilitator to support them in examining the CLDs of their group members and to combine
them into one CLD. A preliminary conceptual model, constructed from the results of the analysis of the interviews, was presented as a
possible starting point. For each modication to the diagram, consensus had to be reached by the group on the proposed adjustment
guided by the facilitators. This activity yielded four CLDs. Then, the process was repeated with the same participants, but we combined
the four groups into two groups (groups 1 and 4 and groups 2 and 3). Each of the new groups then proceeded to combine their CLDs by
discussing the similarities and differences of their two CLDs. This concluded the rst-day session.
After the rst workshop, the research team combined the two CLDs into one and presented it to the group in detail at the beginning
of the second day of the workshop (lasting 2.5 h). A large proportion of the second workshop was spent on discussing, critiquing, and
modifying the aggregated diagram until consensus was reached. At the end of the second day’s session, the participants deliberated on
the impacts of possible interventions using the CLD and expressed their opinions on the overall process. The research team followed up
with an online evaluation survey the next day.
4. Results and discussion
In this section, we describe the summary of the interview transcripts that illustrate stakeholders’ mental maps on the subject
(Section 4.1), the carsharing dynamics model for Bangkok city developed in this study (Section 4.2), the results of the evaluation
survey (Section 4.3), and the reection on the remote GMB (Section 4.4).
4.1. Comparison of stakeholders’ mental models
The interview transcripts were summarized and classied into ve sectors according to the stakeholders’ backgrounds (Table 1).
We highlight and discuss three main ndings here. First, there are apparent differences in how participants in each group visioned a
successful carsharing operation. For example, those from the public sector believed successful carsharing should contribute to public
policy objectives (stimulate GDP, increase employment, and lower transport externalities) and provide a safe and secure (data privacy)
service. In contrast, those from the user group focused on how the service characteristics would meet their needs, while those from the
research group focused on the adoption of carsharing and on certain aspects of the service (easy to use and good coverage) and users at
P. Jittrapirom et al.

Transportation Research Part D 101 (2021) 103122
6
the aggregate level (data privacy). The service providers group was concerned about the economic aspects and in sustaining business
operation (protable and sustaining industry), wider adoption of the service, how service characteristics would be provided, and how
they meet the need of users. Those from the private sector also had an economic concern (self-sustained industry) and also its societal
contributions. The suggested key performance indicators (KPIs) from each group should reect their different focuses but with some
KPIs are in common, such as number of users and reduction of private car sale/ownership.
The diversity of the stakeholders’ views regarding a successful carsharing service may have stemmed from how they perceived the
concept in contributing toward achieving their professional or organizational goals. This observation is similar to Bostrom (2017), who
reported strong correlations between the coherence and consistency of a person’s mental models on a given subject and the person’s
background and expertise. The differences also point toward a potential conict and the importance of aligning these differences as a
shared desirable state to help mobilize and unite efforts among different groups of stakeholders. However, the process would require
some trade-off between competing goals (such as data privacy versus convenience or protability versus sustainability). Although it
was unclear here who should take this leadership and how, the challenges in balancing impartiality, acceptability, and practicability
were apparent and need to be addressed.
Second, the differences among groups in the contributing factors column are less apparent. All the groups believed government
interventions (parking provision, supportive regulation, and policies) are important contributors to carsharing success. All groups also
believed public awareness and users’ understanding of the system to be important. They also suggested several different service-related
factors (easy access, affordable, sufcient eet size). Several groups also believed collaborations among providers and across sectors
would be important. The insights above seem to suggest that the stakeholders can reach agreements concerning these factors (service
characteristics and preferred interventions) more easily. However, this also poses a challenge as stakeholders may avoid having a
challenging discussion concerning any trade-off or alignment of their visions.
Finally, all the policies proposed by all groups, except the providers, were policies focused on beneting carsharing services. Only
one policy was proposed by the provider (restrict private car ownership and use). The supportive policies and regulations include
public transport service development, tax exemption, and parking regulations. Fostering cooperation to support operation imple-
mentation and integration with the public transport service are also among policies suggested by several groups. The lack of ‘push’
policies proposed suggests that interventions that deter personal car use and ownerships are less favorable among the stakeholders,
whereas ‘pull’ policies or policies that promote carsharing and support its operation are more favorable. However, to achieve a
desirable societal goal that is a sustainable transport system, both types of measures would be necessary (Topp and Pharoah, 1994).
4.2. Bangkok carsharing dynamics model
The model developed in this study is based on the diverse and multi-perspective expertise and experience of the participants that
were identied using consensus. It is neither complete nor exhaustive, but it represents the participants’ shared understanding or
aligned mental model of the subject and contains the variables and connections that were deemed the most important by them at the
time. The variables extracted from the interview transcripts were ranked and presented to the workshop participants. Participants
could also suggest additional variables during the workshop. A selection of this list is presented in Table 2, with references to previous
studies; a few of the variables were unique to this study.
The system structure of the model is shown in Fig. 1, Insert A. Several assumptions were made to simplify the diagram; for example,
all net prot is invested to improve the operation and there is an unlimited pool of potential carsharing users to be attracted. The model
shows how the quality of carsharing operation is determined by: a) the quality of the service system and technology, b) the quality of
Table 2
Selected variables from interview transcripts.
Category Variable Reference to previous studies
Vision Convenient and easy to use service (Lane et al., 2015)
Sufcient stations and coverage area (GIZ, 2014)
Range of vehicle types and models n/a
Key performance indicators Awareness of the service (GIZ, 2014)
No. of users or reservations/days, utilization
rate
n/a
Lower transport system externalities (Shaheen et al.,2019)
No. of private cars in the system (GIZ, 2014; Lane et al., 2015; Shaheen et al.,
2019)
Factors that can accelerate or prevent visions from being
reached
Stakeholder collaboration (GIZ, 2014)
Awareness of the service (GIZ, 2014); Lane et al., 2015)
Condence in carsharing service (reliability) n/a
Competitions/alternative modes of transport (GIZ, 2014)
Convenient and attractive services (Lane et al., 2015)
Value for money and customer satisfaction (Lane et al., 2015)
Quality of public transportation systems (GIZ, 2014)
Government support and endorsement (GIZ, 2014)
Government policy Parking regulation (Shaheen et al., 2019)
Tax and road pricing incentive (GIZ, 2014)
Vehicle ownership control policies (Lane et al., 2015)
P. Jittrapirom et al.

Transportation Research Part D 101 (2021) 103122
7
the carsharing eet, and c) the quality of the station and parking; see Table 3 for a detailed description of these attributes. The
identied criteria for carsharing quality are similar to those suggested by Csonka and Csisz´
ar (2016). Additionally, several possible
interventions were also identied and integrated into the CLD. These measures could be from both the public and private sectors.
Three feedback loops that illustrate the dynamics driven by internal attributes within the system were identied:
1. Balancing loop B1: Investment funding inuences the quality of carsharing operation. This balancing loop illustrates that the
quality of the carsharing service is inuenced primarily by investment funding. Each adjustment to an entity within this loop will
result in a new stabilizing level of the other entities. In addition to the investment funding, several public policies (such as dedicated
parking provision, tax incentives, and soft loans) are also important enabling factors of the quality of carsharing operation that
would, in turn, affect the attractiveness of the service and hence the number of users.
2. Balancing loop B2: Service price controls the number of users. This balancing loop follows the traditional economic model in
which carsharing demand is dictated by the price of the service (a stabilizing state). The service price can be lowered through a
reduction in the operation cost or an increase in competition among the operators. On the other hand, the attractiveness of car-
sharing can also be increased if the average cost of other modes increases.
3. Reinforcing loop R1: Service quality attracts users. This reinforcing loop suggests that as service quality continues to increase,
carsharing users will continue to increase exponentially. This system behavior is only hypothetically possible or would be possible
over the short period that the causalities assumed in the model are true. However, in reality, other attributes in the system will
counteract and turn this loop into a balancing loop.
In addition to the three feedback loops identied, Fig. 1, Insert A depicts the dynamics between the government’s recognition
concerning carsharing effectiveness as a sustainable transport measure and its efforts in supporting carsharing implementation. In the
case of Bangkok city, the current carsharing users, public awareness of the system, and the quality of operation are all limited, resulting
in a low level of recognition by the government and hence the marginalized budget and policy efforts allocated to support carsharing.
Three main insights can be drawn from the system diagram. First, collaboration among stakeholders is important for the success of
the carsharing operation. The diagram illustrates how the quality of carsharing and its attractiveness are affected by public policies
(parking regulation, tax incentive, and integration with public transport) and private interventions (investment funding and various
quality contributing factors). Thus, the combined efforts among the stakeholders are essential for the service operation. Second, ten
possible interventions to improve the carsharing operation were identied by the stakeholders. Eight of them are public policies and
measures and two (Promotion of carsharing and Encourage employees to use carsharing) can be considered as private sector policies.
The dominance of public interventions in this list suggests that the stakeholders viewed the public sector as a key enabler. Third, the
‘chicken and egg’ dynamic that limited government efforts to support carsharing (Insert A) would require a public initiative to ‘take the
rst step’ in providing supportive policies for carsharing or employ an adaptive planning approach (Jittrapirom et al., 2018b) in which
its budgets and policy efforts to support carsharing are adjusted according to the outcomes. In contrast, a ‘wait-and-see’ strategy could
Fig. 1. System structure of Bangkok’s carsharing dynamics. Note: The right segment of the gure describes the supply side of the carsharing system
and the left segment the demand side. Suggested policies and measures are depicted in gray boxes and the arrows indicate where these interventions
inuence the system.
P. Jittrapirom et al.

Transportation Research Part D 101 (2021) 103122
8
have a detrimental impact on the carsharing operation.
4.3. Process evaluation by participants
After each workshop, the participants were called during the following days and asked for their feedback. The evaluation survey
showed that the respondents had positive impressions of the workshops (Fig. 2). The participants who responded to the survey (10 out
of 16) appreciated how the workshops enhanced their insights into the problem, gave them a new understanding of the intercon-
nectedness within the system, and brought different perspectives together to reach a shared understanding using the CLD (Q1–6). They
also felt that the workshops were successful and they were satised with the outcomes. (Q7–9). However, they felt that there could be
improvements to ensure that the existing situation and relevant information of the carsharing system are better reected in the CLD
(Q10–11) and more effort should be made regarding the aspects concerning policy analysis (Q12–13). The relatively lower scores on
these questions may be a direct reection of the limited amount of time allocated to address issues related to how the obtained system
structure reects the current situation and how the CLD could support the evaluation of policies and measures.
The participants also commented on and made several suggestions for future activities. Some participants believed the process
helped them to gain insights and appreciate the opinions of others on the subjects (Respondent 1–4 and 9) and helped them to correct
their misunderstanding about carsharing (Respondent 8). Other comments included: the process was easy to engage with and the
facilitators were helpful (Respondents 6,7, and 10); and that CLD is a useful tool to support group communication (Respondents 5,6,
and 10). Additionally, they also reported several drawbacks, such as a drop in interaction level in a certain period (Respondents 4 and
Table 3
Attributes inuencing the quality of carsharing operation identied.
Aspect of quality Contributing attribute
Quality of service system and technology •Tracking and monitoring of vehicle position and condition
•One-way service capability
•Quality of supporting technology (e.g., App, transition, and communication)
Quality of carsharing eet •Fleet size
•Distribution of eet
•Diversity of vehicle size, function, and type
•Cleanliness and functionality
Quality of stations and parking •No. of service pickup points
•No. of temporary parking points
•Station and parking management
•Station and parking distribution
Fig. 2. Results of the online evaluation survey.
P. Jittrapirom et al.

Transportation Research Part D 101 (2021) 103122
9
5), and the need for better briengs on the activity in general (Respondent 5), on CLD (Respondent 1), and on the carsharing system
(Respondents 8 and 9). Finally, several possible improvements were suggested, such as a face-to-face event (Respondent 10) and the
inclusion of high-level decision-makers (Respondent 7) in the process.
The evaluation shows that the GMB process facilitated the stakeholders to share and gain system perspective insights, resulting in
more comprehensive knowledge and a higher acceptance and support for the outcomes systematically (Andersen et al., 1997b;
Rouwette et al., 2011). The application can help to address factors, such as a lack of shared understanding about the concept, that may
slow down or hinder implementations of subjects that may have scare past information or are surrounded by contested outlooks, such
as carsharing in this case. It is also shown to be an efcient method to provide active participation in transport planning and policy
formulating processes in developing countries, such as Thailand, where current transport planning practices are highly ‘top-down’ and
dominated by the public sector with limited stakeholder participation (Jittrapirom and Jaensirisak, 2020). Given the stakeholders’
inuential and unique roles within the system, the shared understand and their empathetical insights are an important basis to support
effective collaboration, support, and consensus-reaching among the stakeholders. It should be emphasized that the resulted model is
not a nal nor a complete representation of the subject; it represents a snapshot of the shared understanding among the participants.
4.4. Reections on remote group model building process
The benets of applying GMB as a methodological approach lie in its transparency and the involvement of stakeholders in both
making and using the model. It takes a systemic approach to problems that cover both technological and social factors (de Gooyert
et al., 2017). The general shortfall of the GMB method lies in its informal approach to elicit knowledge and map causal relationships
within a system (Richardson et al., 1989). Specically, the GMB application in this study assumed that insights into the system could be
constructed through the knowledge and experience of the stakeholders involved. Although the available insights may be sufcient for
the model’s purpose (Forrester, 1987) in establishing a shared understanding, as some key stakeholders were absent, all their per-
spectives may not have been included in the resultant structure. Additionally, the statements and ideas captured were non-explicit,
making it difcult to ‘fact-check’ or determine which of these variables and relationships were more important from the stake-
holders’ perspective. Certain steps within the workshop process could be examined to reveal possible improvements, such as the use of
a preliminary model (Richardson, 2013) and a consensus-reaching process to prevent any potential bias from the halo effect (Nisbett
and Wilson, 1977) or power differences (Vennix, 1996). The program of the workshop itself could also be extended to provide
additional time for policy-related activities.
Another limitation is the underlying assumption of the process; we assumed in this study that better decision-making can be derived
from systematically collect information from stakeholders. This assumption may be acceptable at an early stage of the planning process
but politics, which is intertwined with planning, should be considered more explicitly in the following stages that involved decision
making and trade-offs. Finally, the shared mental model presented here should be considered an initial point of departure for the
stakeholders to continue adjusting their shared mental models in accordance with emerging information and experience gained on the
subject.
Implementing the GMB workshop online also generated new challenges concerning technical and facilitation aspects, several of
which were also observed by Wilkerson et al. (2020). We highlight here some lessons learned from our study. The stakeholder
identication and invitation processes were effective; stakeholders were keen and available to participate. The high levels of will-
ingness and engagement observed may be a possible benet of moving the GMB process online, which should make it easier and more
convenient for stakeholders to join. During the workshops, we observed that not all participants engaged readily in the discussions;
furthermore, the video conference environment in Zoom does not facilitate parallel discussions. The apparent reduction in the
‘richness’ of the communication is a trade-off in using online communication (see, for example, Abrams et al. (2015) for further
discussion on this topic). We also found that during the breakout sessions, it was difcult for the main facilitator to monitor the process
in each breakout room effectively.
5. Conclusion
This study contributes to the implementation of carsharing in Bangkok city and the wider practice by highlighting that stakeholders
connected to the implementation of an innovative mobility service (in our case, carsharing) are likely to have different mental models
about the service, particularly how they vision success for the service. These mental models are likely to be strongly connected to their
backgrounds. The differences suggest the necessity for a process to align their mental models at the initial stage of the policymaking or
planning process to implement such a novel mobility concept.
Additionally, we demonstrated how a remote group model building (GMB) process can address the discrepancies in the mental
models highlighted above. The remote process brought together the relevant stakeholders, who are inuential in the policymaking
process and the operations of the city’s transport system, and facilitated the exchange of knowledge to construct a shared under-
standing among the participants on the new mobility concepts. The process also improved their level of appreciation toward other
stakeholders’ perspectives, aligned their mental models, and enhanced their understanding of the carsharing operation. The devel-
opment of integrated commitments at all levels of governance and the concerted efforts should, in turn, contribute to a successful
implementation of the concept.
The remote GMB process was an effective approach to incorporate stakeholders’ opinions into the transport planning and policy-
making process for the case of carsharing in Bangkok city, where there has been limited opportunity for the public and private
stakeholders to exchange their knowledge together in an active manner. The inclusion of this systematic participation process at the
P. Jittrapirom et al.

Transportation Research Part D 101 (2021) 103122
10
outset of the planning process broadened the perspectives and insights of the subject and this may improve the success rate of
implementation of the transport plan and policy, by helping to avoid misunderstandings and conicts that may delay or derail the
process. Furthermore, the resulting system structure of the process in the form of a causal loop diagram (CLD) was useful in
communicating the initial insights to others and in providing a basis for a formal model in system dynamics to carry out policy analysis
concerning the mobility concept.
The remote setting of GMB was successfully applied in this study, instead of the usual face-to-face setting. The protocol facilitated
participant attendance, reduced operation costs and logistic complications involved in arranging a physical workshop, and critically, it
minimized the risks of exposure to the coronavirus pandemic for all involved.
The lessons learned from the participatory process in this study can be useful for researchers and practitioners seeking to implement
novel mobility concepts, such as Mobility-as-a-Service, to promote sustainability. The systematic GMB process can be applied to align
participants’ understanding of a novel mobility concept and their vision concerning the concept, identifying possible interventions,
and building consensus and commitment in supporting the implementation of the concept.
Future studies could: 1) explore how the knowledge and information generated during the participatory process in GMB can be
utilized to formulate an implementation plan for the subject, 2) examine how a shared mental model changes during stages of planning
and how any trade-off between the stakeholders in a later part of the planning process can be carried out; which stakeholders would be
suitable to take the lead in negotiations; and 3) investigate the application of the GMB approach to support the implementation of other
new mobility services, such as Mobility-as-a-Service, or in supporting a transformative change of the transport system toward a more
sustainable state.
Declaration of Competing Interest
The authors declare that they have no known competing nancial interests or personal relationships that could have appeared to
inuence the work reported in this paper.
Acknowledgements
Support for this research work was provided by the Asian Transportation Research Society (under the research Project Number
2020/005) and their committee. We are grateful to the anonymous reviewers and Dr. Edit Nagy-Tanaka for their invaluable comments
and supports that enabled us to improve the manuscript. The Kasetsart University Research and Development Institute (KURDI),
Bangkok, Thailand provide English editing assistance.
Appendix A. Stakeholders groups and participants involved in the study
Stakeholder group Organization Number of participants Interview Workshop 1/2
1. Policymakers & Public Sectors Transportation Planning and Policy Agency 4 √ √/√
Insurance Regulator 1 √ -/√
Local Authority 1 √ √/√
Land Transport Regulator 4 √ n/a*
Digital Economy Promotion Agency 1 √ √/√
2. Representatives of Users Customer A 1 √ √/√
Customer B 1 √ -/-
3. Research and Smart Mobility experts Smart City Company 1 √ √/√
Smart Mobility Consortium 1 √ √/√
Mobility Researcher A 1 √ √/√
4. Carsharing Providers Operator A 1 √ -/√
Operator B 1 √ √/√
Operator C 1 √ -/-
Operator D 1 √ √/√
5. Private Organizations Property Development 1 √ √/√
Insurance Provider 1 √ √/√
Automaker 1 √ √/√
Total participants 23 23 15/17
* The participants from the land transport regulator declined to attend both workshops.
References
Abrams, K.M., Wang, Z., Song, Y.J., Galindo-Gonzalez, S., 2015. Data Richness Trade-Offs Between Face-to-Face, Online Audiovisual, and Online Text-Only Focus
Groups. Soc. Sci. Comput. Rev. 33 (1), 80–96. https://doi.org/10.1177/0894439313519733.
Ackermann, F., Andersen, D.F., Eden, C., Richardson, G.P., 2010. Using a group decision support system to add value to group model building. Syst. Dyn. Rev. 26 (4),
335–346. https://doi.org/10.1002/sdr.444.
Akkermans, J.A.M., Rouwette, H.A., 1996. Group model-building to facilitate organizational change.
P. Jittrapirom et al.

Transportation Research Part D 101 (2021) 103122
11
Andersen, D.F., Richardson, G.P., 1997. Scripts for group model building. Syst. Dyn. Rev. 13 (2), 107–129. https://doi.org/10.1002/(sici)1099-1727(199722)13:
2<107::aid-sdr120>3.3.co;2-z.
Andersen, D.F., Richardson, G.P., Vennix, J.A.M., 1997. Group model building: Adding more science to the craft. Syst. Dyn. Rev. 13(2), 187–201. https://doi.org/10.
1002/(SICI)1099-1727(199722)13:2<187::AID-SDR124>3.0.CO;2-O.
Andersen, D.F., Richardson, G.P., Vennix, J.A.M., 1997 Group model building: adding more science to the craft - Andersen - 1997 - System Dynamics Review - Wiley
Online Library. https://onlinelibrary.wiley.com/doi/abs/10.1002/%28SICI%291099-1727%28199722%2913%3A2%3C187%3A%3AAID-SDR124%3E3.0.CO%
3B2-O.
ASAP, 2018. ASAP joins Toyota to Provide Carsharing Business. https://www.asapcarrent.com/en/corporate/media/news/94/asap-ผนึกโตโยต้า-หนุนแนวคิด-car-
sharing.
Bangkok Post, 2018. Sansiri adds ride-sharing at projects. https://www.bangkokpost.com/business/1411826/sansiri-adds-ride-sharing-at-projects.
Beutel, M.C., Samsel, C., Mensing, M., Krempels, K.H., 2014. Business model framework to provide heterogeneous mobility services on virtual markets. In:
Proceedings of the 11th International Conference on e-Business, Part of ICETE 2014–11th International Joint Conference on e-Business and Telecommunications,
pp. 145–151. https://doi.org/10.5220/0005118601450151.
Bostrom, A., 2017. Mental Models and Risk Perceptions Related to Climate Change. In: Oxford Research Encyclopedia of Climate Science. Oxford University Press.
https://doi.org/10.1093/acrefore/9780190228620.013.303.
Csonka, B., Csisz´
ar, C., 2016. Service Quality Analysis and Assessment Method for European Carsharing Systems. https://pp.bme.hu/tr/article/view/8559/6976.
CU, 2018. CU -Toyota Launch Ha:mo – Innovative EV Sharing System on Campus. https://www.chula.ac.th/en/news/5552/.
de Gooyert, V., Rouwette, E., van Kranenburg, H., Freeman, E., 2017. Reviewing the role of stakeholders in Operational Research: A stakeholder theory perspective.
Eur. J. Oper. Res. 262 (2), 402–410. https://doi.org/10.1016/J.EJOR.2017.03.079.
Doyle, J., Ford, D., 1988. Mental models concepts for system dynamics research - Doyle - 1998 - System Dynamics Review - Wiley Online Library. https://
onlinelibrary.wiley.com/doi/abs/10.1002/(SICI)1099-1727(199821)14:1%3C3::AID-SDR140%3E3.0.CO;2-K?casa_token=PY34ff-wvuMAAAAA:
UvZLpiSDpHdFr_1yqdV0nVWRRnHQM7SIzyDOd8FeQmCw58sZcrzPjbN-6CXynUnLdURzOfRNAi6jWAZ4.
Esfandabadi, Z., Ravina, M., Diana, M., Zanetti, M.C., 2020. Conceptualizing environmental effects of carsharing services: A system thinking approach. Sci. Total
Environ. 745, 141169. https://doi.org/10.1016/j.scitotenv.2020.141169.
European Commission, 2019. Mobility and Transport Transport in the European Union Current Trends and Issues BACKGROUND INFORMATION.
Forrester, J.W., 1987. Lessons from system dynamics modeling. Syst. Dyn. Rev. 3 (2), 136–149. https://doi.org/10.1002/(ISSN)1099-172710.1002/sdr.v3:210.1002/
sdr.4260030205.
Galletta, A., Cross, W.E., 2013. Mastering the semi-structured interview and beyond: From research design to analysis and publication. In: Mastering the Semi-
Structured Interview and Beyond: From Research Design to Analysis and Publication. https://doi.org/10.5860/choice.51-2430.
Geus, A.de., 1988. Planning as Learning. https://hbr.org/1988/03/planning-as-learning.
Giesel, F., Nobis, C., 2016. The Impact of Carsharing on Car Ownership in German Cities. Transp. Res. Procedia 19, 215–224. https://doi.org/10.1016/j.
trpro.2016.12.082.
GIZ, 2014. Carsharing Services in Emerging Economies. Sustainable Urban Transport Technical Document #12. https://sutp.org/download/7185/. (Accessed 18
October 2021).
Hovmand, P.S. (Ed.), 2014. Community Based System Dynamics. Springer New York, New York, NY.
Jackson, R.B., Friedlingstein, P., Andrew, R.M., Canadell, J.G., Le Qu´
er´
e, C., Peters, G.P., 2019. Persistent fossil fuel growth threatens the Paris Agreement and
planetary health. Environ. Res. Lett. 14 (12), 121001 https://doi.org/10.1088/1748-9326/ab57b3.
Jittrapirom, P., Jaensirisak, S., 2020. A review of Thailand’s transport master plan for regional cities. In: International Review for Spatial Planning and Sustainable
Development, vol. 8 (Issue 2). SPSD Press, pp. 53–69. https://doi.org/10.14246/irspsd.8.2_53.
Jittrapirom, P., Knoacher, H., Mailer, M., 2017. Understanding decision makers’ perceptions of Chiang Mai city’s transport problems an application of Causal Loop
Diagram (CLD) methodology. Transp. Res. Procedia 25, 4438–4453. https://doi.org/10.1016/j.trpro.2017.05.350.
Jittrapirom, P., Marchau, V., van der Heijden, R., Meurs, H., 2018a. Dynamic adaptive policymaking for implementing Mobility-as-a Service (MaaS). Res. Transp. Bus.
Manage. 27 (July), 46–55. https://doi.org/10.1016/j.rtbm.2018.07.001.
Jittrapirom, P., Marchau, V., van der Heijden, R., Meurs, H., 2018b. Dynamic adaptive policymaking for implementing Mobility-as-a Service (MaaS). Res. Transp. Bus.
Manage. 27, 46–55. https://doi.org/10.1016/j.rtbm.2018.07.001.
Jorge, D., Correia, G., 2013. Carsharing systems demand estimation and dened operations: A literature review. Eur. J. Transp. Infrastruct. Res. 13 (3), 201–220.
https://doi.org/10.18757/ejtir.2013.13.3.2999.
Kent, J.L., Dowling, R., 2013. Puncturing automobility? Carsharing practices. J. Transp. Geogr. 32, 86–92. https://doi.org/10.1016/j.jtrangeo.2013.08.014.
Kitzinger, J., 1995. Qualitative Research: Introducing focus groups. BMJ 311 (7000), 299–302. https://doi.org/10.1136/bmj.311.7000.299.
Lane, C., Zeng, H., Dhingra, C., Carrigan, A., 2015. Carsharing A Vehicle for Sustainable Mobility in Emerging Markets? In WRI Ross Center for Sustainable Cities.
http://www.wri.org/sites/default/les/WRI_Report_Carsharing.pdf.
Lane, D., 1992. Modelling as learning: A consultancy methodology for enhancing learning in management teams. Eur. J. Oper. Res. 59 (1), 64–84. https://doi.org/
10.1016/0377-2217(92)90007-V.
Le Pira, M., Inturri, G., Ignaccolo, M., Pluchino, A., 2018. Dealing with the complexity of stakeholder interaction in participatory transport planning. In: Advances in
Intelligent Systems and Computing, vol. 572. Springer Verlag, pp. 55–72. https://doi.org/10.1007/978-3-319-57105-8_3.
Leyden, K. M., Slevin, A., Grey, T., Hynes, M., Frisbaek, F., Silke, R., 2017. Public and Stakeholder Engagement and the Built Environment: a Review. In: Current
environmental health reports, Vol. 4 (Issue 3). Springer, pp. 267–277. https://doi.org/10.1007/s40572-017-0159-7.
Lim, B.C., Klein, K.J., 2006. Team mental models and team performance: A eld study of the effects of team mental model similarity and accuracy. In Journal of
Organizational Behavior, vol. 27 (Issue 4). John Wiley & Sons, Ltd, pp. 403–418. https://doi.org/10.1002/job.387.
Luna-Reyes, L.F., Martinez-Moyano, I.J., Pardo, T.A., Cresswell, A.M., Andersen, D.F., Richardson, G.P., 2006. Anatomy of a group model-building intervention:
Building dynamic theory from case study research. Syst. Dyn. Rev. 22 (4), 291–320. https://doi.org/10.1002/(ISSN)1099-172710.1002/sdr.v22:410.1002/
sdr.349.
Manager Online, 2020. Drivemate provide carsharing service as a preventive measure to Covid-19. https://mgronline.com/motoring/detail/9630000040959.
Marcucci, E., Gatta, V., Le Pira, M., 2018. Gamication design to foster stakeholder engagement and behavior change: An application to urban freight transport.
Transp. Res. Part A: Policy Pract. 118, 119–132. https://doi.org/10.1016/j.tra.2018.08.028.
May, A.D., 2015. Encouraging good practice in the development of Sustainable Urban Mobility Plans. Case Stud. Transp. Policy 3 (1), 3–11. https://doi.org/10.1016/
j.cstp.2014.09.001.
Mingers, J., Sarah, T., 1992. The use of soft systems methodology in practice. J. Operat. Res. Soc. 43 (4), 321–332. https://doi.org/10.1057/jors.1992.47.
MOT, 2017. Digital Transport 2021 Value Creation for Economic & Social Development.
Movmi, 2019. Carsharing Industry: Carsharing Market & Growth 2019. https://movmi.net/carsharing-market-growth-2019/.
Münzel, K., Boon, W., Frenken, K., Vaskelainen, T., 2018. Carsharing business models in Germany: characteristics, success and future prospects. IseB 16 (2), 271–291.
https://doi.org/10.1007/s10257-017-0355-x.
Nisbett, R.E., Wilson, T.D., 1977. The halo effect: Evidence for unconscious alteration of judgments. J. Pers. Soc. Psychol. 35 (4), 250–256. https://doi.org/10.1037/
0022-3514.35.4.250.
Pampel, S.M., Jamson, S.L., Hibberd, D.L., Barnard, Y., 2015. How I reduce fuel consumption: An experimental study on mental models of eco-driving. Transp. Res.
Part C: Emerg. Technol. 58 (PD), 669–680. https://doi.org/10.1016/j.trc.2015.02.005.
Park, A., Suttijaree, J., 2017. Politics of Tourism Promotion : A Case of Car Rent Business in Chiang Mai Province. Polit. Sci. Publ. Admin. J. 2, 81–101.
Senge, Peter M., 1990. The Fifth Discipline. http://kmcenter.rid.go.th/kmc08/km_59/manual_59/Book6/The-Fifth-Discipline.pdf.
P. Jittrapirom et al.

Transportation Research Part D 101 (2021) 103122
12
Purwanto, A., Suˇ
snik, J., Suryadi, F.X., de Fraiture, C., 2019. Using group model building to develop a causal loop mapping of the water-energy-food security nexus in
Karawang Regency, Indonesia. J. Clean. Prod. 240, 118170. https://doi.org/10.1016/j.jclepro.2019.118170.
Quade, E.S., 1972. Analysis for Public Policy Decisions. https://www.rand.org/content/dam/rand/pubs/papers/2008/P4863.pdf.
Rees, D., Stephenson, J., Hopkins, D., Doering, A., 2017. Exploring stability and change in transport systems: combining Delphi and system dynamics approaches.
Transportation 44 (4), 789–805. https://doi.org/10.1007/s11116-016-9677-7.
Richardson, G.P., 2013. Concept models in group model building. Syst. Dyn. Rev. 29 (1), 42–55. https://doi.org/10.1002/sdr.1487.
Richardson, G.P., Vennix, J.A.M., Andersen, D.F., Rohrbaugh, J., Wallace, W.A., 1989. Eliciting Group Knowledge for Model-Building. In: Computer-Based
Management of Complex Systems, pp. 343–357. https://doi.org/10.1007/978-3-642-74946-9_36.
Rouwette, E.A.J.A., Korzilius, H., Vennix, J.A.M., Jacobs, E., 2011. Modeling as persuasion: The impact of group model building on attitudes and behavior. Syst. Dyn.
Rev. 27 (1), 1–21. https://doi.org/10.1002/sdr.441.
Rouwette, E.A.J.A., Vennix, J.A.M., 2019. In: Encyclopedia of Complexity and Systems Science. Springer Berlin Heidelberg, Berlin, Heidelberg, pp. 1–17. https://doi.
org/10.1007/978-3-642-27737-5_264-4.
Rouwette, E.A.J.A., Vennix, J.A.M., Mullekom, T.V., 2002. Group model building effectiveness: A review of assessment studies. Syst. Dyn. Rev. 18 (1), 5–45. https://
doi.org/10.1002/(ISSN)1099-172710.1002/sdr.v18:110.1002/sdr.229.
Ruth, M., Antunes, P., Stave, K., Videira, N., Santos, R., 2015. Using participatory system dynamics in environmental and sustainability dialogues. In: Handbook of
Research methods and Applications in Environmental Studies, pp. 346–374. Edward Elgar Publishing. https://doi.org/10.4337/9781783474646.00022.
Schreier, H., Grimm, C., Kurz, U., Schwieger, B., Kessler, S., M¨
oser, G., 2018. Results of Impact Analysis of Car-Sharing Services and User Behaviour Delivers
Interesting Results in Bremen – Share North. In: Interreg North Sea Region. https://share-north.eu/2018/05/results-of-impact-analysis-of-car-sharing-services-and-
user-behaviour-delivers-interesting-results-in-bremen/.
Scriptapedia, 2020. Scriptapedia - Wikibooks.
Scott, Rodney J., Cavana, Robert Y., Cameron, Donald, 2016. Recent evidence on the effectiveness of group model building. European Journal of Operational Research
249 (3), 908–918. https://doi.org/10.1016/j.ejor.2015.06.078.
Shaheen, S., Cohen, A., Farrar, E., 2019. Carsharing’s impact and future. In: Advances in Transport Policy and Planning, vol. 4 (Issue 1970), pp. 87–120. https://doi.
org/10.1016/bs.atpp.2019.09.002.
Shaheen, S.A., Cohen, A.P., 2007. Growth in worldwide carsharing an international comparison. Transp. Res. Rec. 1992 (1), 81–89. https://doi.org/10.3141/1992-
10.
Shaheen, S., Cohen, A.P., 2013. Carsharing and Personal Vehicle Services: Worldwide Market Developments and Emerging Trends. Int. J. Sustain. Transp. 7 (1), 5–34.
https://doi.org/10.1080/15568318.2012.660103.
Shaheen, S., Martin, E., 2010. Demand for carsharing systems in Beijing, China: An exploratory study. Int. J. Sustain. Transp. 4 (1), 41–55. https://doi.org/10.1080/
15568310802273172.
Spickermann, A., Grienitz, V., Von Der Gracht, H.A., Von Der Gracht, H.A., Von Der Gracht, H.A., 2014. Heading towards a multimodal city of the future: Multi-
stakeholder scenarios for urban mobility. Technol. Forecast. Soc. Chang. 89, 201–221. https://doi.org/10.1016/j.techfore.2013.08.036.
Stephenson, J., Spector, S., Hopkins, D., McCarthy, A., 2018. Deep interventions for a sustainable transport future. Transp. Res. Part D: Transp. Environ. 61, 356–372.
https://doi.org/10.1016/j.trd.2017.06.031.
Sterman, J., 2000. Business dynamics: systems thinking and modeling for a complex world. Irwin/McGraw-Hill.
Strauss, A., Corbin, J.M., 1990. Basics of qualitative research: Grounded theory procedures and techniques. Grounded theory procedures and techniques. Sage
Publications Inc, In Basics of qualitative research.
Topp, H., Pharoah, T., 1994. Car-free city centres. Transportation 21 (3), 231–247. https://doi.org/10.1007/BF01099212.
van Bruggen, A., Nikolic, I., Kwakkel, J., 2019. Modeling with stakeholders for transformative change. Sustainability (Switzerland) 11 (3), 825. https://doi.org/
10.3390/su11030825.
Vennix, J.A.M., 1996. Group model building : facilitating team learning using system dynamics. Wiley.
Voinov, A., Kolagani, N., McCall, M.K., Glynn, P.D., Kragt, M.E., Ostermann, F.O., Pierce, S.A., Ramu, P., 2016. Modelling with stakeholders - Next generation.
Environ. Modell. Software 77, 196–220. https://doi.org/10.1016/j.envsoft.2015.11.016.
Wilkerson, B., Aguiar, A., Gkini, C., Czermainski de Oliveira, I., Lunde Trellevik, L., Kopainsky, B., 2020. Reections on adapting group model building scripts into
online workshops. Syst. Dyn. Rev. sdr.1662. https://doi.org/10.1002/sdr.1662.
Yoshida, T., Yamagata, Y., Chang, S., de Gooyert, V., Seya, H., Murakami, D., Jittrapirom, P., Voulgaris, G., 2020. Spatial modeling and design of smart communities.
In: Urban Systems Design, pp. 199–255. Elsevier. https://doi.org/10.1016/b978-0-12-816055-8.00007-5.
P. Jittrapirom et al.