Discussing Ethical Impacts in Research and Innovation: The Ethics Canvas: 13th IFIP TC 9 International Conference on Human Choice and Computers, HCC13 2018, Held at the 24th IFIP World Computer Congress, WCC 2018, Poznan, Poland, September 19–21, 2018, Proceedings

Abstract

Technologies are increasingly intertwined with people’s daily lives. Consequently, there is an increasing need to consider the ethical impacts that research and innovation (R&I) processes, both in commercial and non-commercial contexts, bring about. However, current methods that offer tools for practicing ethics in R&I inadequately allow for non-ethicists such as engineers and computer scientists to practise ethics in a way that fits the character of their work. As a response, we propose a tool for identifying ethical impacts of R&I that is inspired by a method for the generation of business models, the Business Model Canvas. This tool, the Ethics Canvas, enables researchers to engage with the ethical impacts of their R&I activities in a collaborative manner by discussing different building blocks that together constitute a comprehensive ethical interpretation of a technology. To assess the perceived usefulness of the Ethics Canvas, a classroom experiment was conducted, followed-up by a questionnaire. The results suggest that the Ethics Canvas (1) is perceived as useful for identifying relevant stakeholders and potential ethical impacts and (2) potentially triggers reconsiderations of technology designs or business models.

Full text

This is a draft version of the final manuscript: for the final
paper, refer to “Reijers, W., Koidl, K., Lewis, D.,
Pandit, H.J. and Gordijn, B., 2018, September.
Discussing ethical impacts in research and innovation:
The ethics canvas. In IFIP International Conference
on Human Choice and Computers (pp. 299-313).
Springer, Cham.”
Discussing Ethical Impacts in Research and Innovation:
The Ethics Canvas

1 Introduction
Due to the increasing pervasiveness of technologies in people’s everyday lives (e.g.
social media, artificial intelligence, genomics, communication and transportation
technologies), it becomes increasingly important to reflect on the ethical impacts of
research and innovation (R&I) processes and their outcomes. In academic R&I
settings, ethical considerations are typically mediated by professional codes of
conduct and more or less bureaucratic “ethics clearance” procedures. In commercial
R&I settings, ethical considerations are far less systemised and are often
predominantly taken into account through legal requirements (e.g. in the general data
protection regulation). Both in and outside academia R&I, practices show
considerable shortcomings because they are usually reactive, not suitable for
anticipating potential ethical impacts and not in accordance with the notion of ethics
as a reflective activity. Consequently, there is a need for developing new methods and
tools to guide the practice of ethics in R&I processes, which has led to a burgeoning
literature on practising ethics in R&I.
This paper proposes a novel tool for discussing ethical impacts in the process
of R&I. We conceptualise the R&I process as consisting of four distinct stages: (1)
the formation of (scientific) knowledge and concepts that can be operationalized for
technological applications, (2) the translation of this knowledge into a technology
design, (3) the prototyping and testing of this design and (4) the introduction of the
R&I outcomes into society [1]. Each of these stages of the R&I process can bring
about distinct ethical impacts. However, existing methods that offer tools for
practicing ethics in R&I do not adequately facilitate the discussion of these impacts
amongst R&I practitioners1. Below, we identify an important shortcoming in existing
methods for practicing ethics in R&I. In order to improve the way ethics is currently
dealt with in R&I processes we then propose a novel tool that enables R&I
practitioners to engage with the ethical impacts of their R&I activities.
To achieve this, we searched for an extensively adopted tool that facilitates
accessible and clear processes of discussion and reflection amongst non-experts.
Consequently, we transformed a business-modelling tool that is widely used in
business development practices, the Business Model Canvas (BMC) [2], into the
“Ethics Canvas”. The Ethics Canvas is a collaborative brainstorming tool that has two
distinct aims: (1) to have R&I practitioners come up with and discuss possible ethical
impacts of the technologies they develop, and (2) to have R&I practitioners consider
pivots in their technology design or business model to avoid or mitigate the negative
ethical impacts. The overall aim of this tool is to foster ethically informed technology
design by improving the engagement of R&I practitioners with the ethical impacts of
their R&I activities. We assessed the perceived usefulness of the Ethics Canvas by
putting it into practice in a classroom situation of business & IT students who were
developing novel technological applications. A follow-up questionnaire that the
1 With “practicing ethics” we refer to any way of doing ethics R&I processes in the broadest sense.

students filled in provided some initial suggestions with regards to the usefulness of
the Ethics Canvas for practising ethics in R&I.2
In what follows, we first shortly discuss existing methods that offer tools for
practicing ethics in R&I and discuss an important shortcoming of these methods.
Second, we discuss methods used to create business models that can offer tools for
overcome this shortcoming. Third, we propose the design of the Ethics Canvas that is
inspired by the Business Model Canvas approach in business modelling research.
Fourth, we explore the merits of the Ethics Canvas in a practical setting by evaluating
the outcomes of a questionnaire that students filled in after having engaged in the
Ethics Canvas exercise that related to a technological application they were working
on.
2 Challenges for Practicing Ethics in R&I
In recent years, there has been a strong increase in discussions about responsible R&I
[3], responsible technology design [4] and responsible approaches to instructional
technology research [5]. This tendency is due to growing concerns about ethical
impacts that technological innovations can have on our society, intensifying public
debate and mounting concerns about unsustainable technological developments
(pollution, climate change, etc.). Policy makers reinforce the move towards
responsible R&I, for instance by incorporating ethics assessment practices in funding
mechanisms, as is the case in the framework programmes of the European Union [3].
According to Stilgoe et al. [6], literature on responsible research and innovation
generally focuses on four dimensions of the assessment of R&I practices: (1) the
anticipation of future societal impacts of technology design, (2) reflection on the
values that are implied in technology design, (3) inclusion of stakeholders in the
design process and (4) responsiveness of technology design to societal changes.
Ethical impacts of technologies form an important consideration in this context of
responsible research and innovation, since all of these four dimensions consider
aspects of human-technology interactions or relations that have a strong normative
significance.
Resonating with at least the first three dimensions of responsible R&I,
methods for practising ethics in R&I are (1) dealing with uncertainty of technological
change, such as anticipatory technology ethics [7], (2) enabling, organising and
ensuring ethical technology design, such as ethical impact assessment [8] (3)
identifying, analysing and resolving ethical impacts, such as the ethical matrix [9] and
(4) enabling, organising and ensuring appropriate stakeholder participation, such as
value sensitive design [10, 11]. Not each method for practising ethics falls neatly into
one of these categories of use, but these categories nonetheless provide a useful
overview of the core aspects of practising ethics in R&I that should be taken into
account. In this paper, we will restrict our investigations predominantly to the aspect
of enabling, organising and ensuring technology design.
2 This study has been approved by the Trinity College Dublin Research Ethics Committee

One critical shortcoming of current methods is that they insufficiently
manage to offer tools to integrate ethics in the day-to-day work of R&I practitioners
[11]. Many of the methods for practising ethics in R&I offer tools that are targeted at
ethicists and that presuppose special ethics expertise, which causes difficulties for
R&I practitioners in adopting those tools in their everyday activities. As a result, for
instance, an R&I practitioner cannot simply engage in value sensitive design by
following the proposed steps in the respective literature. Similarly, an ethical impact
assessment needs the involvement of people that manage the assessment process and
offers little grounds for R&I practitioners to assess their day-to-day work.
Yet, the engagement of R&I practitioners in practising ethics is crucial for
achieving ethical technology design. In the overall process of practising ethics, Brey
[12] distinguishes a disclosure level, at which ethical impacts are explored and
identified, a theoretical level, at which theoretical frameworks are developed and
employed to evaluate these impacts and the application level, at which moral
deliberation takes place as the basis for overcoming the negative ethical impacts in the
R&I process. According to Brey, the disclosure level is aimed at revealing, or rather
identifying the potential ethical impacts of a technological application. He accentuates
the role of R&I practitioners in dealing with ethics in R&I at this level, arguing that
they play a vital role in disclosing ethical issues of emerging technologies and in
making sure that technology design choices are informed by considerations of these
ethical issues [12]. Brey argues that researchers, designers and innovators are
important actors at the disclosure level, because they have an understanding of the
technology that many ethicists and policy-makers lack. Hence, we have good reasons
to look for a novel tool that overcomes the shortcomings of tools offered by existing
methods and facilitates engagement of R&I practitioners with ethics in their day-to-
day work3.
Resulting from the foregoing considerations, we propose two requirements
that such a tool should satisfy. First, it should be accessible to non-ethicists, or people
without a substantial background in ethics. This argument is reinforced by the Council
for Big Data, Ethics and Society, which argues that ethics engagement should happen
in “hybrid spaces” in which people with different roles in the R&I process collaborate
[14]. Second, it should enable people with different roles and backgrounds to work
together in identifying ethical impacts. The different interpretations people have of
potential ethical impacts of technologies can be brought to the table by means of a
collaborative process in which multiple people involved in an R&I process express
their expectations of potential ethical impacts in the form of narratives (i.e. for this
group of stakeholders, such-and-such feature of our technology can have such-and-
such ethical implications). As we will see later, this requirement fits with theories in
Science and Technology Studies (STS) that explain how our understanding of
technologies results from an interaction between different interpretations of
technologies for different people [15].
3 For our current purposes and the scope of this paper it is not feasible to propose a full-fledged method.
However, the Ethics Canvas tool fits with a newly developed method that introduces a narrative approach
in ethics of technology [13].

In line with these two requirements, we have to make sure that the threshold for
using our tool should be low and that it should be possible to use it without thorough
background knowledge of ethical theories or conceptual discussions of values.
Additionally, the tool should facilitate an open-ended process of interpretation in a
collaborative fashion to identify potential ethical impacts of an R&I process and its
outcomes. At the moment, methods in the fields of applied ethics and ethics of
technology offer no tools that adequately fulfil these two requirements. For this
reason, we decided to look at fields that are unrelated to academic ethics, but that do
focus on creating low-entry tools for collaborative processes of discussion and
interpretation; focusing on a specific use case. This brought us to the field of business
development, and notably the field of business model development.
3 Turn to the Business Model Canvas
In the field of business model development, some discussions of responsible research
and innovation have already emerged. For instance, Henriksen et al. discuss business
models that promote sustainable ways of production, or “green business model
innovation” [16]. In a similar vein, Bocken et al. explain how a re-definition of the
notion of “value” in business models can help rendering businesses more sustainable,
for instance by focusing on the entire supply-chain [17]. However, these approaches
primarily focus on the design of the business case (in terms of resources, customers,
etc.) and not on technologies that are developed in R&I processes. To change this
focus, we will investigate how we can transform existing business model
development approaches to align them with our aim of constructing a tool for
disclosing the ethical impacts in R&I processes in which technologies play a crucial
role.
Before we do so, however, we have to address the question of to what extent
tools in business model development methods incorporate the two requirements we
introduced in the previous section: (1) engaging non-ethicists with the disclosure of
ethical impacts and (2) facilitating this as a collaborative process of interpretation. We
can observe how business models are defined in the literature. As pointed out by Zott
et al. [18], a business model can be understood as an “architecture”, a “heuristic
logic”, a “concise representation” and also a collection of “stories”, aimed at
describing and explaining how a particular enterprise functions or operates. A
spectrum of business models can be identified, with on the one hand business models
that are meant to offer a strict representation of both internal and external processes of
an existing corporation, and on the other hand models that use stories to give an
account of these processes – possibly of businesses that do not already exist (i.e. a
model for a start-up). The latter type of business modelling approaches is particularly
interesting for our purposes, since it appears to focus on an understanding of business
processes in terms of narratives that are constructed through social interaction.
Lucassen et al. [19] use two indicators that capture the two aspects of the
abovementioned spectrum (between models that are strictly representational and those
that are the result of people’s interpretations) to review and compare different visual

business modelling approaches. They use the notion of “capturing” to indicate to what
extent a business modelling method accurately represents a business process, and the
notion of “communicating” to indicate to what extent a business-modelling tool is
accessible and generates understandable outcomes. They argue that the so-called
Business Model Canvas (BMC) is most successful with regards to the indicator of
“communicating”, compared to two other established models4 “because it effectively
models explicit information of both tangible and intangible aspects of a business and
communicates this information in a highly accessible manner to parties unfamiliar
with the modelling technique” [19]. As Kuparinen argues, the BMC can be classified
as a “narrative business model” [20], because it enables “participant narratives” [21].
The BMC provides a visual-linguistic tool (see figure 1) that can be used in a
collaborative process in which participants generate ideas by offering and discussing
certain narratives that are related to the thematic boxes displayed on the canvas.
Fig. 1: The BMC [2].
Thus, we argue that of the existing business modelling approaches, the BMC fulfils
the two requirements we formulated. First, as Lucassen et al. [19] argue, it is highly
accessible and understandable to people without specific knowledge of the field. If the
structure of the BMC can be incorporated in a tool for disclosing ethical issues, it
would be an answer to Brey’s [12] concern regarding the disclosure level for it allows
researchers to engage with ethical reflection in an accessible manner without them
having to have thorough knowledge of the field of applied ethics. Second, since the
BMC relies on the collaborative generation of participant narratives, it seems to
4 The two traditional business model approaches that the business model canvas was compared with are
the “software ecosystem model” approach and the “board of innovation” approach [19].

satisfy the second demand to a large extent. It enables participants to engage in a
collaborative process of interpreting and discussing business processes. Considering
the foregoing arguments, turning towards the BMC to find a novel tool for disclosing
ethical impacts in R&I processes is justified. However, we need to transform the
BMC, which is clearly focused on discussing business processes and has little to do
with ethics, into a tool that can be used in the context of practicing ethics in R&I.
4 Designing the Ethics Canvas
To explain the process of designing the Ethics Canvas, we first briefly describe the
BMC and discuss its aims and the way in which it is used in a collaborative setting.
The BMC was developed by business theorists Osterwalder and Pigneur [2] as a
visual-textual plane that is divided up into nine “building blocks” through which a
business model can be described in a holistic manner. It is argued that a business
model can be defined as a model that “describes the rationale of how an organisation
creates, delivers, and captures value” [2], and that this definition can be captured by
participants discussing all the “building blocks” of a business model. By engaging in
a collaborative discussion about the different building blocks of a business model,
such as ‘key partners’, ‘channels’ and ‘revenue streams’, participants working with
the BMC are able to arrive at a comprehensive understanding of the way in which
their organisation is supposed to create, deliver and capture value.
In its original form, the BMC is printed on paper and used as the core
instrument for a collaborative workshop. In addition to the canvas, Osterwalder and
Pigneur provide for a handbook that provides guidance for the workshop participants
in understanding the meaning of the different building blocks and presents use-cases
of the canvas as well as techniques for designing better business models. The printed
canvas is used as the focal point of a collaborative workshop, with participants
discussing and writing down ideas for each of the building blocks. Next to the original
BMC, online applications have been developed that offer digital versions5 of the
canvas, through which teams of different sizes can create multiple business models
and save them on their accounts. The BMC has experienced widespread adoption in
the business modelling of start-ups. Its ease of use in capturing and communication a
business model lends itself well to the identification and resolution of uncertainties
typically facing teams developing a start business model. Blank [22] describes how
BMC is widely used in the teaching of start-up business modelling to research and
innovation across universities in the United States. In this capacity, the BMC is used
as an easily updated ‘scorecard’ for documenting the development of pivots in a
business model when following Blank’s own iterative, evidence-driven Customer
Development methodology [23] combined with Reis’ agile approach to start-up
development [24]. Osterwalder et al. [25] have elaborated how the value proposition
and customer segments elements of the BMC can be further categorised in the
detailed modelling and testing of product market fit. These developments indicate that
5 See for instance https://strategyzer.com/ and https://canvanizer.com/new/business-model-canvas

the form of the BMC has proven adaptable both to integration into independently
developed methodologies as well as for methodological elaboration in critical areas.
Taking the business model canvas as a starting point, we aimed to transform
it in a way that would enable its users to discuss how a technology might bring about
ethical impacts for different stakeholders instead of discussing a business model. To
achieve this, we considered different building blocks that could amount to a holistic
ethical analysis of a certain technological application. The building blocks were
constructed in a two-way process: by considering literature in Science, Technology
and Society studies (STS) and philosophy of technology, and simultaneously
engaging in a trial-and-error exercise of using the Ethics Canvas to improve its user-
friendliness. Each building block consists of a central term and a number of core
questions that can guide the discussion concerning a term.
We consulted literature that provides accounts of different aspects of impacts
of technologies on individuals, groups and society as a whole. We need to stress that
this consultation was mostly aimed at pragmatically gathering different vantage points
to consider ethical impacts of technology, and not at providing a coherent theoretical
framework underpinning the Ethics Canvas. The STS literature offers useful accounts
of the ways in which technologies are embedded in relationships between different
“relevant social groups” [15], which can be types of individuals (e.g. producers,
technology users, women, elderly) or institutional, collective actors (e.g. government,
companies, labour unions). Akrich [26] discusses the Actor-Network Theory (ANT)
approach and shows how technologies can have impacts on actors that are not directly
connected to its design, production or use such as non-users but also non-humans
(understood as e.g. the impact of a mobile phone on a supply chain for raw materials).
She argues that technologies can politicise social and material relations, which can for
instance be made explicit by considering how non-users of social media applications
can become marginalised.
In order to subsequently understand how technologies impact relevant
individuals or groups, we turn to writings in philosophy and technology. Ihde [27] and
Verbeek [28] show how individuals can change their behaviour or relationships by
engaging with technologies. For instance, Verbeek shows how the ultrasound
technology has transformed the relationship between parents and their unborn child
[28], and how technologies, such as traffic lights and speed bumps, mediate the
behaviour of car drivers [28]. These scholars accentuate that “ethical impacts” are not
simply consequences of technological change, but should be understood as impactful
relations between human beings and technologies. Feenberg [29] goes beyond this
focus on the technological mediation at the level of the individual, by arguing how
technologies can impact relations between people and collectives, for instance
between workers and their companies, between governments and labour unions. In
line with this, he shows how technologies can impact the public sphere, in which
“everyday communicative interactions” take place [29], in which ideologies are
formed and social struggles arise. To consider ethical impacts that are more directly
related to the material aspect of technologies, we consulted scholars discussing
“constructive technology assessment”. These show that technology assessment should
take impacts of technologies on the environment and production processes into

account [30]. Finally, to provide a bridging step in the move from description (i.e.
what are the ethical impacts?) to prescription (i.e. what should be done?), we turn
once more to value sensitive design and included the notion of technical choices
driven by value considerations [10] as the logical end-point of the Ethics Canvas.
However, we broadened up the choices to be considered, asking participations to
think beyond the technical by also considering organisational changes or changes in
policies.
Although we did not provide a full-fledged and exhaustive review of literature
dealing with the impacts of technologies on humans, groups, and society as a whole,
our discussion does give us an adequate picture of what the building blocks of the
Ethics Canvas could look like. To summarise, we can infer the following
characterisations of technology impacts from the literature:
• Ethical impacts occur as relations between technologies and different
types of actors, which can be types of individuals and types of collectives,
or groups.
• Technologies can mediate the behaviour of individuals, but also the
relations that people have with one-another.
• Technologies can mediate the worldviews of social groups and can bring
about social conflicts between social groups.
• A technology impacts the material network in which it is designed,
produced and used, including for instance the supply chain it constitutes
and the natural resources it needs.
While taking the abovementioned characterisations of technology impacts as a
guideline, we entered into a trial-and-error design process of the Ethics Canvas. This
design process was aimed at making sure that the rationale of the Ethics Canvas
design would not only be grounded in the relevant literature, but that it would also be
user-friendly and intuitive. Based on the literature, we designed nine different
versions of the Ethics Canvas, all with different building blocks and layouts. These
designs were iterated through a series analysis exercises conducted by the Ethics
Canvas design team, which consisted of the authors of this paper and other
researchers who collectively possessed expertise in applied ethics, personalisation in
digital applications, knowledge engineering, software engineering and innovation
methodologies. In addition, versions of the Ethics Canvas were trialled in teaching
and training settings with over 260 undergraduate and postgraduate students in
computer science, engineering, business studies and working on groups on pre-
assigned digital application designs. This provided a further source of design insight
into improving the usability of the Ethics Canvas design. The criteria for success we
used during these meetings were that participants (1) should be able to complete the
entire canvas within a reasonable amount of time (a maximum of 1,5 hours) and (2)
should be able to address each building block without having to consult any external
source.
As a result of this trial-and-error exercise, some important changes were
made concerning the wordings of each box, because some terms use in the consulted
literature (e.g. actor, human-technology-world relations, ideology) were not intuitive
for the users and needed to be translated into concepts that are more easily usable (e.g.

group, behaviour, worldview). The table below (table 1) provides an overview of the
conceptual framework of the Ethics Canvas, displaying sources in the academic
literature and the corresponding approaches that each building block is based on and
explicating what changes in terminology were applied to ensure the usability of the
Ethics Canvas.
Table 1: Overview of (1) the central questions of the Ethics Canvas, (2) authors consulted
to address these, (3) the approaches used by these authors, (4) the changes in wordings applied
during the trial-and-error sessions and (5) the final boxes for the Ethics Canvas.
Central
question
Literature
consulted
Approach
Change in wording
Boxes
Who are
affected?
Pinch and
Bijker [15],
Akrich [26]
Actor Network
Theory
Relevant social
group / actor /
actant =>
individual / group
1) Individuals
affected
2) Groups
affected
How are
stakeholders
affected?
Ihde [27],
Verbeek
[28]
Postpheno-
menology
“Human-
technology-world”
relation =>
behaviour /
relations
3) Behaviour
4) Relations
Feenberg
[29]
Critical
Theory of
Technology
Ideology =>
worldviews
Struggles => social
conflicts
5) Worldviews
6) Social
Conflicts
Schot and
Rip [30]
Constructive
Technology
Assessment
Risks of products
and processes =>
product or service
failure
Environmental
aspects =>
Problematic use of
resources
7) Product or
service
failure
8) Problematic
use of
resources
What can be
done?
Friedman,
Kahn, and
Borning [10]
Value
Sensitive
Design
Technical choices
driven by value-
considerations =>
What can we do?
9) What can we
do?
Eventually, the design process brought us to the current design of the Ethics
Canvas (figure 2). The Ethics Canvas is organised according to nine thematic blocks
that are grouped together according to four different stages of completing the canvas.
The first stage (blocks 1 and 2) challenges the participants to consider which types of
individuals and groups might be relevant stakeholders when considering a specific
technology. The second stage (blocks 3 to 6) asks the participants to consider
potential ethical impacts, considering the different stakeholders that were identified.

The third stage (blocks 7 and 8) asks the participants to consider potential ethical
impacts that are not stakeholder specific, pertaining to product or service failure or
any problematic use of resources. The fourth stage (block 9) challenges participants to
think beyond the potential ethical impacts they discussed and discuss some initial
ideas for overcoming these ethical impacts. To complete the Ethics Canvas exercise in
a physical space, participants can write down their ideas on a printed Ethics Canvas,
and consult the Ethics Canvas Manual [31] that provides guidance on how to conduct
the exercise. An online version of the Ethics Canvas6 has also been developed. On
this platform, people can collaborate to complete a particular Ethics Canvas online
while being in different physical places.
Fig. 2: The Ethics Canvas, version 1.9.
5 Assessing the Usefulness of the Ethics Canvas
The BMC is a widely used tool for business model development and has been
positively assessed [19]. We wanted to similarly assess the Ethics Canvas and its
usefulness as a tool that supports practising ethics in R&I settings. Comparing the
Ethics Canvas with other tools for practising ethics in R&I is not possible due to lack
of similar tools that are used in day-to-day activities of R&I practitioners. Therefore,
6 https://ethicscanvas.org/index.html

we assessed the Ethics Canvas through evaluating its perceived usefulness amongst its
users and its anticipated effects related to follow-up activities.
We organised an Ethics Canvas pilot with students who were required to
create a new ICT application as part of their coursework. The students attended a one-
hour lecture at which the content of the Ethics Canvas Manual was presented. After
this, the students were given the assignment to complete the Ethics Canvas in groups,
using the online version for their particular R&I project in approximately one hour.
Students were free to meet up in a physical space or to hold a conference call for
completing the exercise. A total of 109 students participated in the Ethics Canvas
exercise, organised into groups, each comprising of 3 or 4 students. After the groups
had completed their Ethics Canvasses, all participating students were asked to fill in a
questionnaire that asked them about their perception of the usefulness of the Ethics
Canvas to practise ethics in their respective R&I projects. Filling in the questionnaire
was voluntary. The feedback questionnaire was filled in by 31 students, which
represented 28% of the total number of students who worked on the Ethics Canvas
exercise.
The questionnaire followed a 5-point Likert scale, with a 1-point assessment
indicating strong disagreement and a 5-points assessment indicating strong
agreement. Statements were formulated in the affirmative mode and as negations to
be able to assess whether participants paid attention to the statements. The
participants were asked about (1) the perceived usefulness of the Ethics Canvas (e.g.
did the Ethics Canvas add to the overall understanding of ethical considerations?) and
(2) the anticipated effect of the Ethics Canvas (e.g. did the exercise influence the
business model and or technology design?). In what follows, these two aspects are
discussed based on reflections on the questionnaire results.
The perceived usefulness of the Ethics Canvas was evaluated extensively in
the questionnaire. Generally, 56% of the participants agreed and 28% strongly agreed
that the exercise improved their understanding of the potential ethical impacts of their
R&I projects. Participants were asked whether the Ethics Canvas exercise widened
their understanding of different individuals or groups affected by their project, to
which 44% of the participants replied that they agreed and 29% that they strongly
agreed. On being asked whether the exercise helped them create a broad overview of
potential ethical impacts of their project, 42% of the participants stated to agree and
35% to strongly agree. To further the scope of the assessment, the participants were
asked whether the ethical impacts they discussed in the task sufficiently fitted the
structure of the Ethics Canvas. 40% of the participants agreed that it sufficiently fitted
and 21% strongly agreed. To assess the value of the Ethics Canvas in stimulating
productive discussions, participants were asked whether they considered any ethical
impacts that were not known to them or unclear beforehand. Only 21% of the
participants disagreed or disagreed strongly with this question, indicating that the
majority of the participants discussed ethical impacts that were new to them. This
suggests that the Ethics Canvas can be a useful tool to guide participants into
discussing ethical impacts that group members didn’t know or didn’t clearly think
about beforehand.

The second theme of the survey focused on the assessing the anticipated effect
the Ethics Canvas has the business model and technology design of the ICT
application that the students are working on. First the participants were asked whether
the exercise would have any impact on their project’s technology design, resulting in
32% of the participants agreeing and 16% strongly agreeing. A similar question was
asked in relation to the impact of the canvas on the business model. 52% of the
participants agreed that the Ethics Canvas led them to reconsider their business
models and 5% strongly agreed. Finally, the participants were asked whether the
exercise was useful in promoting the group’s ethical behaviour. 35% of the
participants agreed that the exercise promoted ethical behaviour and 40% even
strongly agreed. Even though these outcomes do not directly indicate that follow-up
actions have been taken or will be taken, they at least indicate an intention amongst
the students to use the outcomes of the Ethics Canvas exercise to adjust their business
models or technology designs.
Overall, the results suggest that it is reasonable to state that the Ethics Canvas is
perceived as a useful tool to guide participants in discussing a broad range of ethical
impacts as well as the identification of relevant stakeholders. Moreover, the results
indicate that it is reasonable to assume that the Ethics Canvas can lead to the intention
of participants to reconsider their business models or technology designs.
Nevertheless, results also indicate that the structure of the Ethics Canvas will need to
be improved to be more inclusive of potential ethical impacts. Moreover, our study is
limited due to the limited participation rate (28% of all the students who worked on
the Ethics Canvas exercise). This might possible have led to biased results, because
the cohort of students that voluntarily filled in the questionnaire could have coincided
with the cohort of students that was most positively engaged during the Ethics Canvas
exercise. Hence, even though these initial results positively suggest that the Ethics
Canvas is a useful tool for practising ethics in R&I, further development of the Ethics
Canvas and additional ways of assessing its usefulness will be needed for future
studies.
6 Conclusion
In this paper, we presented a novel tool for the discussion of ethical impacts in R&I
settings. The Ethics Canvas responds to shortcomings in the current methods that
offer tools for practicing ethics in R&I, concerning the lack of ways in which R&I
practitioners can engage in practising ethics in their day-to-day work. We argued for
two requirements for constructing a suitable tool for addressing the disclosure stage:
that it should be accessible and clear to R&I practitioners and that it should facilitate a
collaborative process in which people can discuss different interpretations of impacts
of technologies. Since no existing tool in the field of ethics of technology seemed to
address this need in an adequate way, we turned to the field of business model
development instead. In this field, we assessed the BMC as a suitable tool because it
is highly accessible to different types of people, and structures a collaborative effort
to discuss issues surrounding a central goal. We designed the Ethics Canvas by re-

directing the focus of the canvas format from business modelling to a comprehensive
identification of ethical issues of an R&I process. Utilising established theories in
philosophy of technology and STS that are aimed at understanding ways in which
technologies can impact the behaviour and relations of individuals and collectives,
and engaging in a trial-and-error design process, we formulated different building
blocks of the Ethics Canvas. Finally, we put the Ethics Canvas to the test in a
classroom setting, which resulted in initial positive results, which suggest that the
Ethics Canvas is perceived as a useful tool for discussing relevant stakeholders and
potential ethical impacts in R&I projects and for triggering anticipations of pivots in
business models or technology designs. However, more studies will need to be done
to further develop the Ethics Canvas and assess its usefulness in multiple ways.
Finally, we should reflect on two limitations of the Ethics Canvas that could
prompt future research. First, even though the exercise can provide R&I teams with a
much better overview and understanding of ethical impacts of their R&I activities it
does not yet provide a way to evaluate these impacts. That is, is does not provide a
way to evaluate whether a certain ethical impact is to be considered positive or
negative or whether it is to be considered severe or non-severe. Additional tools will
thus have to be developed to enable this, which will probably have to draw strongly
from theories in normative ethics (i.e. consequentialism, deontology, virtue ethics)
that offer the best intellectual resources for shaping evaluation practices. Second, the
Ethics Canvas draws from conceptual work in academic literature, but does not yet
offer a way to translate engagement with the Canvas in practice to revisit its
conceptual roots. Potentially, multiple Ethics Canvas exercises could for instance be
used as empirical input for revisiting and refining the actor network theory. Future
work could therefore focus on the translation of concrete and numerous outcomes of
Ethics Canvas exercises into changes in the conceptual framework that guide our
thinking about R&I activities.
Acknowledgements. The ADAPT Centre for Digital Content Technology is funded
under the SFI Research Centres Programme (Grant 13/RC/2106) and is co-funded
under the European Regional Development Fund. In addition, this paper has received
funding from the European Union’s Horizon 2020 research and innovation
programme under grant agreement No 700540.
References
1. Roberts, E. B.: Managing invention and innovation. Research Technology
Management, 50(1), 35–54 (2007)
2. Osterwalder, A., & Pigneur, Y.: Business model generation: a handbook for
visionaries, game changers, and challengers. Retrieved at December 1 2017 from
https://www.businessmodelgeneration.com (2010)
3. Owen, R., Macnaghten, P., Stilgoe, J.: Responsible research and innovation: From
science in society to science for society, with society. Science and Public Policy,
39(6), 751–760 (2012)

4. Tromp, N., Hekkert, P., Verbeek, P.-P.: Design for Socially Responsible Behavior: A
Classification of Influence Based on Intended User Experience. Design Issues, 27(3),
3–19 (2011)
5. Reeves, T., Herrington, J., Oliver, R.: Design research: A socially responsible
approach to instructional technology research in higher education. Journal of
Computing in Higher Education, 16(2), 97–116 (2005)
6. Stilgoe, J., Owen, R., Macnaghten, P.: Developing a framework for responsible
innovation. Research Policy, 42(9), 1568–1580 (2013)
7. Brey, P.: Anticipatory Ethics for Emerging Technologies. NanoEthics, 6(1), 1–13
(2012)
8. Wright, D.: Ethical impact assessment. Ethics, Science, Technology, and
Engineering, 163(c), 163–167 (2014)
9. Forsberg, E.: The Ethical Matrix — A Tool for Ethical Assessments of
Biotechnology. Global Bioethics, 17(1) (2004)
10. Friedman, B., Kahn Jr., P. H., Borning, A.: Value Sensitive Design and Information
Systems. In K. E. Himma & H. T. Tavani (eds.), Human-Computer Interaction and
Management Information Systems: Foundations (pp. 1–27) Hoboken, New Jersey:
John Wiley & Sons, Inc. (2006)
11. Reijers, W., Wright, D., Brey, P., Weber, K., Rodrigues, R., O’Sullivan, D., Gordijn,
B.: Methods for Practising Ethics in Research and Innovation: A Literature Review,
Critical Analysis and Recommendations. Science and Engineering Ethics (2017)
12. Brey, P.: Disclosive Computer Ethics: The Exposure and Evaluation of Embedded
Normativity in Computer Technology. Computers and Society, 30(4), 10-16 (2000)
13. Reijers, W., Gordijn, B., O’Sullivan, D. Narrative ethics of personalisation
technologies. In D. Kreps, G. Fletcher, & M. Griffiths (Eds.), IFIP Advances in
Information and Communication Technology (Vol. 474, pp. 130–140). Cham:
Springer International Publishing. (2016)
14. Metcalf, J., Keller, E. F., Boyd, D.: Perspectives on Big Data, Ethics, and Society.
The Council for Big Data, Ethics, And Society. Retrieved on December 1 2017 from
http://bdes.datasociety.net/council-output/perspectives-
on-big-data-ethics-and-society/ (2006)
15. Pinch, T. J., Bijker, W. E.: The Social Construction of Facts and Artifacts: or How
the Sociology of Science and the Sociology of Technology Might Benefit Each
Other. Social Studies of Science, 14(3), 221–232 (1984)
16. Henriksen, K., Bjerre, M., Bisgaard, T., Høgenhaven, C., Almasi, A., Grann, E.
Green Business Model Innovation: Empirical and Literature Studies. Nordic
Innovation Report. Nordic Innovation: Oslo (2012)
17. Bocken, N., Short, S., Rana, P., Evans, S.: A value mapping tool for sustainable
business modelling. Corporate Governance: The International Journal of Effective
Board Performance, 13(5), 482–497 (2013)
18. Zott, C., Amit, R., Massa, L.: The business model: Recent developments and future
research. Journal of Management, 37(4), 1019-1042 (2011)
19. Lucassen, G., Brinkkemper, S., Jansen, S., Handoyo, E.: Comparison of Visual
Business Modeling Techniques for Software Companies. In: Cusumano, M., Iyer, B.,
Venkatraman, N. (eds.) Software Business: Third International Conference, ICSOB

2012, pp. 14. Cambridge, MA, USA, June 18-20, 2012. Proceedings (2012)
20. Kuparinen, P.: Business Model Renewal and its Networking Aspects in a Telecom
Service Company. (Master Thesis). Tampere: Tampere University of Technology
(2012)
21. Oliveira, M., Ferreira, J.: Book review: Business model generation: A handbook for
visionaries, game changers and challengers. African Journal of Business
Management, 5(7) (2011)
22. Blank, S.: Why the Lean Start-Up Changes Everything, Harvard Business Review,
91(5), 63-72 (2013)
23. Blank, S.: The four steps to the epiphany: successful strategies for products that win.
Pennsauken, New Jersey: BookBaby (2013)
24. Reis, E.: The Lean Startup. Lake Arbor: Crown Publishing (2011)
25. Osterwalder, A., Pigneur, Y., Bernarda, G., Smaith, J.: Value Proposition Design.
Hoboken, New Jersey: John Wiley & Sons (2014)
26. Akrich, M.: The De-scription of Technical Objects. In W. Bijker, J. Law (Eds.),
Shaping technology, Building society. Cambridge, Massachusetts: MIT Press (1992)
27. Ihde, D. Postphenomenology and Technoscience. New York: Sunny Press (2009)
28. Verbeek, P.-P. What things do; philosophical reflections on technology, agency, and
design. Pennsylvania: Pennsylvania University Press (2005)
29. Feenberg, A. Questioning Technology. New York: Routledge (1999)
30. Schot, J., Rip, A. The past and future of constructive technology assessment.
Technological Forecasting and Social Change, 54, 251–268 (1997)
31. Lewis, D., Reijers, W., Pandit, H. Ethics Canvas Manual. Retrieved on December 20
2017, from https://ethicscanvas.org/download/handbook.pdf
(2017)