Ethical Perception of a Digital Study Assistant: Student Survey of
Ethical Values
Paul Greiff
a
, Carla Reinken
b
and Uwe Hoppe
c
Institute of Information Management and Information Systems Engineering, Osnabrück University, Osnabrück, Germany
Keywords: Digital Study Assistant, Ethical Values, Students.
Abstract: The digital transformation in higher education progresses constantly. Here, new technical innovations are
emerging, such as a digital study assistant (DSA). The DSA is designed to help students to identify and
achieve their personal study goals. In this regard, it should be noted that ethical considerations play an
increasingly important role in the introduction of digital systems and thus also in the DSA. Therefore, the
user-centered perspective is taken into account in the development of a DSA by addressing personal ethical
values. For this purpose, two consecutive surveys were conducted with 42 and 156 students from a German
university. The aim of the work is to identify ethical values in relation to the DSA that were perceived as
particularly important by students as the main user group. From this, practical implications and further
research possibilities regarding DSAs and ethical issues can be derived.
1 INTRODUCTION
Progressive digitalization has made a major impact on
society in the twenty-first century. The way in which
people communicate, exchange information, develop
and understand disciplinary knowledge has changed
dramatically with the development and availability of
digital technologies (Ihme and Senkbeil, 2017). As a
result, digital innovations have developed far-
reaching effects on our moral life and thus on current
ethical issues of digitalization that need to be
addressed (Floridi, 2010). This is where our study
comes in, as we want to take a closer look at the
ethical perception of a digital study assistant in a
higher education context from a student's point of
view. In doing so, the interests of a heterogeneous
student body must be given special consideration
(Allemann-Ghionda, 2014).
In addition to digitalization, the academic
landscape has been shaped by the internationalization
of study structures, the increasing permeability of the
education system, and the pluralization of lifestyles
(Zervakis and Mooraj, 2014). Decision-makers are
therefore confronted more than ever with the question
of the direction in which universities should develop
a
https://orcid.org/0000-0002-8159-946X
b
https://orcid.org/0000-0001-7295-6368
c
https://orcid.org/0000-0002-9186-1468
in order to meet new challenges (Heuchemer, 2018).
To support students efficiently and effectively in
achieving their individual educational goals, the
development of virtual assistants or so-called digital
study assistants (DSA) has therefore become
increasingly important (Alexander et al., 2019).
The development, implementation, and
evaluation of such a DSA have been taking place
since November 2018 within the framework of the
joint project SIDDATA. The digital assistant is
intended to support students in their actions based on
a situation analysis and give them recommendations
for achieving predefined goals. Such a digital study
assistant can help to realize a wide range of potentials,
both on the institutional side and on the students' side.
Academic institutions can better understand the
learning needs of their students and positively
influence their learning and their learning progress
(Slade and Prinsloo, 2013). The choice of modules
and periods of study abroad can be made easier for
students by providing information in line with their
interests. In addition, it has been shown that chat
offers, for example, can be used as an autonomous
learning instrument (Benotti et al., 2014; Dutta, 2017;
Abbasi and Kazi, 2014). The basis for this is
92
Greiff, P., Reinken, C. and Hoppe, U.
Ethical Perception of a Digital Study Assistant: Student Survey of Ethical Values.
DOI: 10.5220/0011063800003182
In Proceedings of the 14th International Conference on Computer Supported Education (CSEDU 2022) - Volume 1, pages 92-104
ISBN: 978-989-758-562-3; ISSN: 2184-5026
Copyright
c
2022 by SCITEPRESS Science and Technology Publications, Lda. All rights reserved
comprehensive data access from students, through
which the systems can provide decision support
according to user preferences. It seems to be valuable
that a better understanding of the student body has a
number of advantages. However, the collection and
use of personal data can lead to various moral,
political, and economic dilemmas (Munoko et al.,
2020). It is therefore essential to address issues of
privacy, security, self-determination, and justice at an
early stage (Manzeschke, 2020). In the field of digital
technologies, different ethical concepts can be found,
depending on the research topic, which partly
overlaps, like computer ethics as a separate part of
technical ethics (Johnson and Miller, 2009; Moor,
1985). Further ethical areas in this field are machine
ethics (Anderson and Anderson, 2011), robot ethics
(Lin and Abney, 2017), and information ethics
(Floridi, 2015). Although ethics is regarded by many
experts as an integral part of technology assessment,
there is a great need for further research in this
context, for example, for measures of ethical attitudes
or frameworks for ethical impact assessment (Wright,
2011; Masrom et al., 2010; Harris et al., 2011; Millar,
2016). Therefore, the aim of this study is to identify
ethical values that are considered important for
students as the central stakeholder group for the DSA.
For this reason, the following research questions are
posed:
RQ1: Which ethical values are considered
important by students in relation to a DSA?
RQ2: What correlations can be found between the
ethical values mentioned by students?
2 THEORETICAL FOUNDATION
2.1 Digital Study Assistants as Part of
Digital Transformation
As a result of the constantly growing opportunities to
use digital innovations, the level of digitalization at
universities is also increasing analogously.
Technological progress in recent years made it
possible to bundle a large amount of student data.
Students have access to a wide variety of digital
resources, are increasingly networking online, and are
interacting more and more on a wide variety of digital
platforms (Ihme and Senkbeil, 2017). In order to be
able to use student data to their advantage,
technologies such as assistance systems (e.g. DSA)
and learning analysis are becoming more and more
important for the future development of universities.
They are associated with a number of positive effects
for students, professors, and the universities. As
Rouse already pointed out, technological changes are
closely related to transformation (Rouse, 2005). The
term digital transformation (DT) conquers the
modern world and describes the use of new digital
technologies to enable major improvements
(Fitzgerald et al., 2013). These technologies are not
new per se, it is often more about the combinations
and evolving possibilities that create a new
innovation like it is the case with the SIDDATA
project. DT is regarded as a major change in society
and business and is often described as an ongoing
process (Morakanyane, 2017). DT is an important and
contemporary issue in academic education and cannot
be neglected in the context of a digital study assistant
(Gottburgsen and Wilige, 2018). Changing learning
conditions in the age of digitalization must be
perceived for further implementation in order to
interact dynamically and flexibly (Ahel and
Lingenau, 2020). New technologies in higher
education require a certain level of user acceptance in
order to be able to sustainably survive on the market
and above all to guarantee long-term added value for
students and other stakeholders (Mukerjee, 2014).
Various challenges like the internationalization of
study structures or the increasing permeability of the
education systems (Zervakis and Mooraj, 2014) are
putting academic institutions under great pressure.
Traditional approaches must be reconsidered and
replaced or supplemented by new ideas. It is therefore
important that higher education institutions are
supported by the academic community in the
development of new business models and the
implementation of innovation (Hold et al., 2017).
In recent years, the development of digital
assistance systems in particular has gained enormous
importance in the field of business informatics, this is
shown in the latest NMC Horizon Report. The NMC
Horizon Report from 2014 and 2019 lists virtual
assistants as one of six important future technologies
in the context of higher education (Alexander et al.,
2019; Johnson et al., 2014). This refers especially to
cognitive assistance systems with regard to the
provision of information and communication. These
serve above all to provide application-oriented
information in work and learning processes (Apt et
al., 2018). The aim of DSAs is to support students in
their actions through a situation analysis and to give
them recommendations for achieving predefined
goals. In digitalization, however, there are more
extensive possibilities and potential uses for the
development of such systems. Central capabilities of
digital assistance systems at the current state of
research are environmental perception, reactive
behavior, attention control, and situation
Ethical Perception of a Digital Study Assistant: Student Survey of Ethical Values
93
interpretation. In the future, assistance systems
should offer adaptive, situational, and individualized
support using sensory detection of the user and
context (Apt et al., 2018). A DSA could, for example,
react to requests from learners and support students in
their everyday study routine. Such a system could
support staff in advising and informing students and
teachers with specific didactic and organizational
tasks. Students could be supported in the self-
organization of their studies in the form of a
"reflection partner" (Schmohl and Schäffer, 2019).
The project SIDDATA seeks to examine whether
and how students can be efficiently and effectively
assisted in achieving individual educational goals by
bringing together previously unrelated data and
information in an individual DSA. The use of the
DSA is intended to encourage students to define and
consistently pursue their own educational goals. In
the future, the data-driven environment should be
able to provide situation-appropriate hints, reminders,
and recommendations, including local as well as
externally offered courses and Open Educational
Resources (OER). In this project, in addition to the
development of the mentioned functions, ethical
considerations also play a key role in order to meet
the requirements of the students. The DSA is initially
implemented and evaluated at three universities.
Students should be encouraged to define and
consistently pursue their own study objectives, and to
be supported by a data-driven environment. The
implementation of a DSA requires technical
guidelines at the strategic level for a structured
approach by universities to adapt to these changes
(Leal et al., 2020). It is also important to consider user
acceptance, e.g. through consideration of ethical
aspects, to ensure sustainable use by students,
teachers, and employees of organizational
departments of a university (Hirsch-Kreinsen et al.,
2015).
2.2 Ethics in Digital Technologies
Due to the progressing digitalization in higher
education, the question is becoming more relevant
according to which moral and ethical standards digital
technologies are developed and used. For this reason,
the investigation of moral and ethical norms or
phenomena in digitalization, even a separate ethics
branch, the information and computer ethics, was
established. According to Pardon and Siemens, ethics
in the digital context can be defined “as the
systematization of correct and incorrect behavior in
virtual spaces according to all stakeholders” (Pardon
and Siemens, 2014). Concerns about moral tensions
(Willies, 2014) and ethical dilemmas have been
raised in the past. These are associated with the
processes of data collection, data mining, and
learning analytics implementation (Drachsler et al.,
2014; Shum and Ferguson, 2012).
In the research and development of human-
technology interaction, ethical aspects are often
considered insufficiently or too late (Brandenburg et
al., 2018). At the same time, the research,
development, and use of innovative technologies
have always required ethically responsible action
from all stakeholders (Ropohl, 1996). Recent
thinking about ethics of information technology (IT)
and computer science has therefore focused on how
to develop pragmatic methodologies and frameworks.
These assist in making moral and ethical values
integral parts of research and development and
innovation processes at a stage in which they can still
make a difference. These approaches seek to broaden
the criteria for judging the quality of IT to include a
range of moral and human values and ethical
considerations. Moral values and moral
considerations are construed as requirements for
design. This interest in the ethical design of IT arises
at a point in time where we are at a crossroad of two
developments: first, “a value turn in engineering
design” and on the other hand “a design turn in
thinking about values” (van den Hoven, 2017). It is
assumed that technology is not value-neutral. Value-
Sensitive Design (VSD) recognizes that the design of
technologies bears “directly and systematically on the
realization, or suppression, of particular
configurations of social, ethical, and political values”
(Flanagan et al., 2008).
The adoption and entry into force of the General
Data Protection Regulation of the European Union is
a current example of how the protection of personal
data and the right to informational self-determination
play an important role in regulations and public
debates. In the further development of innovative
technologies, ethical values should therefore also be
anticipated at an early stage and taken into account in
the design (Brandenburg et al., 2017). New forms of
data analysis, including machine learning, have
greatly increased the effectiveness and speed of data
analysis in recent years. According to the British
Academy and Royal Society, these aspects build the
foundation that renders an approach for the use of
data indispensable. This foundation represents a key
factor for broad acceptance and is therefore an
important building block for the success of digital
innovations (British Academy, 2018). During the
development of a DSA, it is particularly essential to
consider the ethical values from the perspective of the
CSEDU 2022 - 14th International Conference on Computer Supported Education
94
students, as acceptance should be high especially
among this stakeholder group. According to the VSD
approach to ethics of technology, ethical analysis and
moral deliberation should not be construed as abstract
and relatively isolated exercises resulting in
considerations situated at a great distance from
science and technology. Instead, VSD should be
utilized at the early stages of the research and
development (van den Hoven, 2017). Therefore, this
paper focuses on the identification of relevant ethical
values from the user's perspective in order to
incorporate them into the development process of the
DSA at an early stage.
3 RESEARCH DESIGN
For this paper, two separately conducted studies were
included. The content of these two studies is based on
each other, with the results of study 1 being integrated
into study 2. First, an exploratory survey was
conducted with students (n=42). This survey contains
questions about ethical drivers and barriers regarding
a potential use of a DSA from the students'
perspective. The results of study 1 have a dual
function. On the one hand, they already directly
depict a result of which ethical values are important
to students regarding their use or non-use of a DSA.
On the other hand, the results were used as a basis to
develop categories which were used for a quantitative
survey (study 2, n=156). Figure 1 schematically
illustrates the procedure used here.
Figure 1: Research Design.
In the next sub-sections, the individual method-
logical approaches of both studies will be discussed
in detail.
3.1 Method – Study 1
An explorative, qualitative short questionnaire in
online form was created, which a total of 42 students
completed in full. This method was chosen because it
is fundamental in an exploratory procedure to ask
opinions and expectations of the participants freely
and as unbiased as possible. This survey mode is
preferable to an interview approach in the current
pandemic situation and at the same time can be
carried out independently of the participants' time.
The questionnaire started with a welcome text, which
also explained the aim of the survey. This was
followed by an informational text about the
characteristics and goals of a DSA. This information
was followed by questions about what the DSA
should fulfill from an ethical point of view in order to
use it and what ethical barriers would lead students
not to use the DSA. The survey addressed students at
a German university and for this reason, the survey
was conducted in German. Table 1 shows the
structure of the questionnaire with the corresponding
questions.
Table 1: Structure of the Questionnaire.
Question
group
Question
Answer
mode
Information text about the DSA
Ethical drivers
What would be
particularly important to
you in terms of ethics
for long-term use of a
digital study assistant?
Free text
Ethical barriers
What ethical barriers
would be prohibitive for
you to use a digital
study assistant?
Free text
Demographics
Please indicate your
gender.
How old are you?
What field of study are
you currently studying?
Drop-
down
Number
Free text
Since the questionnaire has a strongly qualitative
character due to the free-text answers, a qualitative
analysis method was used to evaluate the results.
Here, a procedure was chosen that is oriented towards
qualitative content analysis (Mayring, 2015). The
chosen procedure is divided into four phases. First,
the answers were sorted by question and paraphrased
(if necessary). Then, the paraphrases were
generalized to core sentences at an appropriate level
of abstraction (phase 2). In the third phase, the first
reduction was made by shortening semantically
identical core sentences and those that were not
Ethical Perception of a Digital Study Assistant: Student Survey of Ethical Values
95
perceived as contributing significantly to the content.
Finally, as the second reduction, the core sentences
were combined with similar or identical sentences
and thus classified into categories (phase 4).
The sample consisted of students from a German
university. Of the 42 respondents, 24 participants
classified themselves as female and 15 as male. One
respondent stated being diverse and two respondents
did not provide any information regarding gender. In
terms of the age group of the sample, it was found that
nine participants were under 20 years old, in the age
group 20-24 there were 18, from 25-29 there were
eleven and four participants were over 30 years old.
Students from different fields of study also
participated in the survey. Students of social sciences
were the most represented with 13 participants,
followed by education students with eight
respondents and students of economics with five
participants. Furthermore, natural sciences (four
participants), computer science (three participants)
and administrative sciences (two participants) were
represented. Five participants were assigned to other
courses of study and two respondents made no
statement in this regard.
The aim of this survey in the first step was to
develop ethical value categories that can serve as an
indicator of what is important to students from an
ethical point of view. In the second step, the
categories collected serve as the basis for the
subsequent quantitative survey.
3.2 Method – Study 2
In order to investigate the ethical values collected
from Study 1, a quantitative questionnaire with items,
which represent ethical statements, was developed.
Consequently, the respondents move within a
predefined grid of answer options. In this case, a six-
point Likert scale is used. The response options range
from “- - - do not agree at all" to “+++ fully agree".
Since the main user group of the DSA are students,
the survey is exclusively addressed to enrolled
students from a German university, like in study 1.
For this reason, the survey was also conducted in
German. The aim of this survey was to evaluate the
identified ethical values from study 1 in terms of their
relevance and importance by the students.
In the beginning, the participants are given an info
text on the topic of the survey and motivation. In this
context, the students had the opportunity to watch an
image film for a better understanding of the DSA and
the SIDDATA project. Before starting the
questionnaire, the students were given detailed
information regarding the goal and content of the
DSA, since they could not be provided with a version
of the DSA yet. This should ensure that the students
develop an idea about the DSA and that ethical
implications arise for them. The questionnaire
comprises a total of 15 questions, with the first
question being an example question. This example
question was intended to familiarize students with the
usage of the Likert scale in the survey. For the
development of the questionnaire the most named
ethical value categories from the qualitative survey
were used. Here, for each ethical values, five items
were formulated in the first step. Some of the items
have been formulated in such a way that they are
negatively polarized in order to avoid response
patterns. Subsequently, a focus group consisting of
six researchers was assigned and reduced the items.
The purpose was to select items that best represent the
corresponding category of the ethical value (e.g.,
fairness). This procedure left three items for five
categories. Of these remaining items, five have
negative polarity. Afterwards, a pre-test was
conducted with ten students to check and adjust the
comprehensibility and wording of the items.
Participants should express their agreement or
disagreement with the items by stating their own
opinion using the Likert scale. These opinions can
provide information about which ethical values,
already mentioned in the first study, are also
perceived as relevant from the perspective of students
in relation to a digital study assistant. Table 2 shows
the three Likert items for the fairness category as an
example.
Table 2: Items of the Ethical Value Category Fairness.
Question
group
Question Polarity
Fairness
I think fairness towards the
users of a digital study
assistant is elementary.
Positive
I would not care if the DSA
favored or disadvantaged
certain groups of people.
Negative
If I perceive the DSA to be
unfair, then that would be a
reason for me not to use the
system.
Positive
The questionnaire took an average of
approximately 10 minutes to complete, including
reading through the info text and watching the image
video. The survey was conducted digitally through
the survey tool LimeSurvey (www.limesurvey.org).
A total of 227 enrolled students of the Osnabrück
University participated in the survey. Of these, 156
CSEDU 2022 - 14th International Conference on Computer Supported Education
96
students completed the questionnaire in full. 71
students partially skipped questions or abandoned the
survey prematurely. Since the demographic data in
Study 1 did not reveal any relevant differences, they
were not considered in Study 2. The survey was
addressed to all enrolled students at the Osnabrück
University and was not limited to a specific semester
or department.
Finally, the data were evaluated and analyzed
using the statistical program SPSS. Since the Likert
scale used in this context does not contain any metric
data, it is important for the further processing of the
data in SPSS that the response options are
transformed. Since the answer "do not agree at all" is
a clear statement of complete disagreement, this
statement is equated to 0. The other answer options
are then rated in ascending order, so that "fully agree"
is equated with the highest value of 5. For the
negatively worded items, the results were then
reversed so that a fully agree (5) equals 0 and a do not
agree at all (0) equals a 5. This ensures that the results
are presented correctly. SPSS was used to create a
reliability analysis, the collection of descriptive
statistics and inter-correlations.
4 RESULTS
4.1 Results – Study 1
With regard to the drivers that favor long-term use of
a DSA, the students clearly see the topic of data
protection in the lead (Table 3). 21 core sentences
were assigned to the category of data protection,
putting this category clearly at the top of the rankings
ahead of second place. It is important to the students
that their data is not passed on and that the collection
of data by the DSA is tied to a specific purpose and
that this purpose is not subsequently extended. A
privacy by design approach was also suggested in this
context, in order to take data protection into account
as early as the development stage. Transparency was
taken up in a total of nine key phrases. In this
category, the respondents emphasized that it is
important to them how their data is handled and also
how and by which algorithms the study assistant
arrives at its results or generates information.
In third place, with seven core sentences each, are
autonomy and data security. The category autonomy
is described by the students as control over functions,
information control and the possibility of being able
to decide as freely or autonomously as possible. Data
security is distinguished from data protection in these
categories in that it describes protection against
external attack or intrusion. Data protection, on the
other hand, primarily describes protection against the
transfer of data outside the system. The last rank is
fairness, in this case with five core sentences. This
outlines that the DSA should not favor or discriminate
against anyone and should be available to all students
for free use.
Table 3: Ethical Drivers.
Question
group
Selection of
mentioned
core sentences
Number of
assigned
core
sentences
Category
Ethical
drivers
for usage
Protected content
to which only
selected individuals
have access;
privacy by design;
the study assistant
should not share
the data; protection
of individual data.
21 Data privacy
Transparency of
how the collected
data is used;
transparency and
consent when the
DSA proposes
something,
publishing of the
program code.
9
Transparency/
Informed
consent
User control over
functions; own
influence on
selection and
presentation of
information;
independent
decision making.
7 Autonomy
Securing data
against loss and
third-party access;
protection against
hacking; high data
security.
7 Data security
No preference in
proposals; no
discrimination;
opportunity for use
by all students.
5 Fairness
After the drivers, the barriers are considered next
in Table 4. Data privacy, which took first place
among drivers, is now also represented in first place
among barriers, with 18 core sentences. The students
surveyed considered the greatest barrier to using the
Ethical Perception of a Digital Study Assistant: Student Survey of Ethical Values
97
DSA to be the disclosure of personal data or even
uncertainty about this issue. They clearly stated here
that lack of privacy would be a strong criterion for not
using the study assistant.
Table 4: Ethical Barriers.
Question
group
Selection of
mentioned
core
sentences
Number of
assigned
core
sentences
Category
Ethical
barriers
against
usage
Disclosure of
personalized data
to third parties;
uncertainty that
own data would not
be handled
properly, data
privacy concerns;
no purpose
limitation of data..
18
Lack of
data
privacy
No freedom of
decision; no
sufficient control;
autonomous
assumptions of the
system;
optimization to
norm study time.
10
Violation
of
autonomy
Request of too
much personal
data; no anonymity
given;
accumulation of
personal data.
6
Lack of
(data)
anonymity
Possibility to use
not given to all
students; have a
lead that non-users
don't have.
4
Unfair-
ness
System could be
hacked; lack of
data security
4
Lack of
data
security
In second place with ten core sentences is the
category violation of autonomy. According to the
students surveyed, a lack of freedom to make
decisions, not having sufficient control, or feeling
forced into a role would be a barrier to use. Lack of
anonymity ranks third with six core sentences.
According to the respondents, this relates to the
request for too much personal data or when
anonymity should not be given. Fourth place among
the barriers is shared by the categories unfairness and
insufficient data security. A barrier to use is seen
when the DSA acts unfairly, i.e. users have an
advantage over non-users or not all students can/are
allowed to use it. Another barrier seen by students is
insufficient data security, which could, for example,
lead to the DSA not being able to withstand an
external attack. The following ethical value
categories, which were derived from the drivers and
barriers serve as the basis, for the second survey in
study 2. Data Privacy/Anonymity: Due to a great
overlap in the students' statements, the categories of
drivers and barriers of data privacy and the barrier
lack of (data) anonymity were merged. A distinction
between the two categories was not expected by the
students. In today's information age, privacy is one of
the main concerns in society and research (Johann
and Maalej, 2013). Privacy is understood as the
ability and/or the (legal) right of an individual person
or group to seclude themselves or information about
them from third parties. With regard to the protection
of information privacy, this means that personal data
is secured against unauthorized access (data security)
and also that only an authorized group of people is
granted access to this data (data privacy) (Ienca et al.,
2017). Fairness: Particularly concerning digital
inclusion, this category represents a core value for
ensuring that as many people as possible from
different backgrounds can participate in and use
digital technologies (Kernaghan, 2014). Fairness here
means the equal distribution of opportunities, rights,
goods through technology and equal access to a
technology (Ienca et al., 2017; Steinmann et al., 2015).
Autonomy: This ethical value refers to the possibility
(in this case through technology) that people are free
to decide, plan and act as they wish in order to achieve
self-determined goals (Friedmann et al., 2013). The
term autonomy also often refers to self-determination.
Related to the ethical context of DSA, this means that
students are granted the opportunity to act in a self-
determined and autonomous manner (Keber and
Bachmeier, 2019). This includes freedom through
third-party monitoring, supervision, and
categorization (Cohen, 2000). Data Security: (Data)
Security refers to protection against destruction or
theft of information structures and data by
unauthorized third parties. It is often referred to as IT
security, computer security, and information security
(Gasser, 1988). Transparency/Informed consent:
Transparency here refers to the disclosure and
communication of functions and ways of data
processing of the DSA. Informed consent refers to the
consent of students to the use of their (personal) data,
including its revocation. It should be noted that
comprehensive information about the nature and use
of the data must be provided beforehand (Keber and
Bachmeier, 2019). Consent must be given voluntarily
CSEDU 2022 - 14th International Conference on Computer Supported Education
98
after the person has been informed of the possible
effects and risks. If possible, this consent should be
given in text form or by a clear statement of consent
(Wright, 2011).
4.2 Results – Study 2
In the first step, the quality criteria of the
questionnaire are explained before the descriptive
results and the correlations are discussed. To ensure
the content validity of the questionnaire, the focus
group was first used to assign and reduce items. The
subsequent pre-test with students also contributes to
ensuring that the understanding of the items is as
consistent as possible, thus ensuring inter-
subjectivity. For the reliability analysis in form of an
internal consistency test, Cronbach’s alpha was
calculated. The internal consistency of a Cronbach’s
alpha = .84 can be considered as satisfying.
First, the descriptive findings are examined and
classified with regard to the first research question.
As mentioned above, after closing the survey, we
transformed the results to obtain metric data for
calculation. Accordingly, the highest achievable
value for the agreement of the ethical statements
represents 5 and the lowest is 0. The transition from a
single minus (-) to a single plus (+) is seen as the level
at which the students agree with the thesis at least to
a small extent. Consequently, a mean value of 2.6
represents the lowest possible level of agreement. The
standard deviation (SD) for the ethical value
categories is between 0.7 and 0.9, indicating a low
degree of dispersion. Table 5 shows the mean values
and the associated standard deviations for the ethical
value categories, which are discussed below.
Furthermore, the table presents the ranking of the
categories from study 1 for comparison. First of all, it
should be noted that there was a level of agreement
on all categories of ethical values by the students. As
described above, this agreement would already have
been reached with a mean value of 2.6. However,
since each category reached at least a value of M =
3.5, it can be assumed that the students as a whole
attach a certain relevance to them.
Fairness: The respondents in the survey consider
the category fairness (M = 4.3, SD = 0.8) to be the
most important ethical value. Thus, students see the
fairness of the digital study assistant as the most
important factor with regard to the consideration of
ethical values. Access to the DSA should be made
equally available to all students, including all user
groups, and treat them equally. In addition,
respondents also explicitly stated in one item that a
perceived unfairness of the system would lead
students not to use it. Transparency/Informed
consent: This ethical value category is close behind
with a mean of 4.2 and a standard deviation of 0.9
comes in second place as an important ethical value.
Here, students expect the DSA to inform them in
detail about the use and processing of their provided
data. Furthermore, no data should be used or shared
for any purpose other than that declared without the
explicit consent of the user. The third rank is shared
by three ethical value categories with a mean value of
4.0. These can thus still be interpreted as important
ethical values with regard to the research question.
Table 5: Comparison Study 1 and Study 2.
Results Study 2 Results Study 1
Ethical
value
category
M SD Characteristic Rank
Fairness 4.3 0.8 Driver
Barrier
4
th
4
th
Trans-
parency/
Informed
consent
4.2 0.9 Driver 2
nd
Data
Privacy/
Anony-
mity
4.0 0.9 Driver
Barriers
1
st
1
st
; 3
rd
*
Data
Security
4.0 0.9 Driver
Barrier
3
rd
4
th
Autonomy 3.5 0.9 Driver
Barrier
3
rd
3
rd
n = 156, M = mean value, SD = standard deviation
*separate rankings before category were merged
Data Privacy/Anonymity: The category data
privacy/anonymity achieved a mean value of 4.0 with
a standard deviation of 0.9. With this result,
respondents confirm that privacy is highly important
to them in a digital study assistant and that it would
be a criterion for non-use if the DSA did not respect
their privacy. Students would also care about the
purposes for which their data would be used within
the system. Data Security: In the same line, (data)
security also achieved a mean value of 4.0 and a
standard deviation of 9.0. Here, students indicated
that (data) security is a high priority for them and if
they had security concerns with the DSA, they would
not share data with the system. Respondents also
indicated that the issue of data security was not
overrated within the context of a DSA. The following
categories safety, accountability, and autonomy did
Ethical Perception of a Digital Study Assistant: Student Survey of Ethical Values
99
Table 6: Correlations of the Ethical Value Categories.
Fairness Transparency/
Informed consent
Data Privacy/
Anonymity
Data Security
Fairness
Transparency/Informed consent .35
Data Privacy/Anonymity .33 .56
Data Security .34 .64 .71
Autonomy .46 .33 .27 .37
not reach the necessary minimum value of M =
4.0 to be considered important but are nevertheless
briefly examined below. Autonomy: The lowest
mean value in this survey was reached by autonomy
with M= 3.5, which can nevertheless still be
evaluated as clear agreement due to the Importance of
this ethical value.
The students agree that they are given freedom to
make personal decisions in planning their studies, for
example. In the course of evaluating the results,
correlations of the ethical value categories were also
carried out. Table 3 above shows the correlations of
the ethical value categories, with the high correlations
(Pearson) above .50 shown in bold. The highest inter-
correlation found was between privacy and (data)
security with .71. This result suggests that privacy
and (data) security are considered very similarly by
the students surveyed, meaning that a clear line
between these two categories may not be valid.
It might be useful, also in terms of item reduction,
to merge these two categories or try to formulate them
more distinctly in the future. The second highest
correlation between the categories of ethical values
was found between (data) security and informed
consent (64). In this case, as well, it can be assumed
that there is at least a partial overlap between the two
categories. The situation is similar with the privacy
and informed consent categories. Here, the inter-
correlation of the two categories is .56. It seems that
there is thus a triangular relationship between the
three categories privacy, (data) security, and
informed consent. It was already noted in the focus
group and the pre-test that these are in fact quite
similar, but that there is a clear distinction between
these categories. There was also a correlation of .46
between fairness and autonomy. A possible
explanation for this could be that autonomy could
pick up on a partial aspect of fairness. Here it could
be useful to specifically look for connections between
the contents of these two categories.
5 DISCUSSION
The results show that all the ethical categories
surveyed are attributed a certain importance by the
students. Especially with the second study, these
categories should be differentiated with regard to
their importance. However, it can be stated here that
at most marginal differences were found, which
makes it difficult to assess the most important ethical
value categories. In addition, the categories all
achieved at least a mean value of 3.5 (autonomy),
which is equivalent to a range between + and ++ on
the Likert scale. Therefore, all underlying categories
are considered important for the use of a DSA from
the student's point of view. Although, as mentioned
above, it is difficult to provide a clear hierarchy of the
importance of the ethical values, the significance of
the individual results will be discussed below. The
results point out that four of the five value categories
appear particularly important to the students, as these
have a mean score of 4.0 and higher.
Here, the fairness of the DSA represents a
fundamental ethical value from the perspective of the
students surveyed. This study showed that students do
not accept that the DSA is perceived as unfair and that
this can lead to non-use. In the first study, fairness
was mentioned both as a driver and in negative form
as unfairness as a barrier. In both cases, the fourth
rank was reached in accordance to core sentence
mentions. In the second study, fairness was ranked
first with a mean value of 4.3. It is thus interpretable
that fairness is perceived as more important if it is
explicitly named as an ethical value in advance. In
contrast, fairness seems to play a less important role
when students reflect unbiased about drivers and
barriers of an DSA. To address fairness, DSA
developers could consider in preliminary stages the
areas in which fairness conflicts may arise. It is
important to identify exactly what is perceived as
unfair and to take preventive measures accordingly.
CSEDU 2022 - 14th International Conference on Computer Supported Education
100
The open questions in the first study pointed out, for
example, different treatment of different groups of
people are seen as unfair. Here, too, it is not yet clear
exactly what characteristics (e.g. gender, nationality
or course of study) can be linked to this. One way to
address for example nationality, in the interest of
fairness, is to design a DSA in a multilingual fashion.
Thus, foreign students have an equal understanding
of functions and recommendations of a DSA and can
therefore use it more effectively.
Respondents also have a clear opinion regarding
transparency and informed consent. They want
transparency and also be informed about the use of
their data and expect that this also sets the framework
for actual data use. Students also see it as an important
factor that they give informed consent for the use of
their personal data. The results show that
transparency and consent are also perceived as
inextricably linked by the respondents. This category
was exclusively named as a driver in study one and
was represented here in second place. In the second
study, transparency/informed consent received a
mean value of 4.2 and thus also achieved second
place. For developers and operators of DSAs, it is
therefore important to clearly communicate the use of
the data and also to obtain the consent of the user
group in advance. If possible, it could also be
considered to make the corresponding source code
publicly available to create maximum transparency
and traceability of the DSA.
In this context, it can be noted that data
privacy/anonymity also has an important role to play
with regard to personal data. Data privacy ranks first
in study 1 for both barriers and drivers, making it the
most important ethical value for students in relation
to a DSA in this case. Within the qualitative content
analysis in study 1 it was found that the core sentences
of insufficient data privacy and the lack of (data)
anonymity have great overlaps. Thus, for study 2, the
values of data privacy/anonymity were combined. In
the second study, a mean value of 4.0 is subsequently
achieved. Concerns about violating the data
privacy/anonymity of study assistant users could be
addressed in several ways. Data Privacy governs how
data is collected, shared and used. Students clearly
express the concern that their data could be shared
with third parties and used for other purposes as
stated. In this context it became obvious, that this
category overlaps with data security and
transparency/informed consent, which is also
highlighted in the correlations. Persons responsible
for the DSA should receive regular training on data
privacy so that they understand the processes and
procedures required to ensure the proper collection,
sharing and use of sensitive data as part of a general
data management portfolio. The data management
portfolio plays a crucial role not only in the data
privacy/anonymity category, but also in the data
security category. When developing a digital study
assistant, care should be taken to preserve the
anonymity of the students. Therefore, those
responsible for the DSA should clarify which data is
really important so that the DSA can be used
effectively. Identifying characteristics which are not
necessary should be negated from the data sets in this
context to ensure the desired anonymity of the
students. A similar situation occurs with the data
security category. Data security ranks third among
drivers and fourth among barriers as an ethical value
for using a DSA. Insecure systems, hacking, and fear
of losing one's data were particularly highlighted by
students in study 1. In the second study, data security
is also rated as very relevant with a mean value of 4.0.
To ensure high standards of data security, data
protection measures and access controls must be in
place to ensure that only those with the appropriate
access rights can view the data. Likewise, steps must
be taken to protect the data from loss or destruction,
for example through regular data backups or a
firewall to protect against external access. In this
context, the creation of a detailed data security
concept according to University policies and the
current law also plays a central role in preventing
hacker attacks.
Autonomy was ranked third as an ethic driver and
barrier in study 1. In the second study, this category
dropped noticeably compared to the others, achieving
only a mean score of 3.5. Here it can be assumed that
autonomy do not seem to be of great importance to
the students. One possible explanation is, that
students are willing to sacrifice part of their autonomy
in order to receive advice from the study assistant,
even if this is perceived as patronizing. It is also
interesting to look at the individual items of
autonomy. Respondents are more likely to agree on
the importance of autonomy than on the consequence
of not using the DSA if their autonomy is restricted.
This result should be interpreted cautiously, however,
as a mean of 3.5 can still be considered a clear
agreement on the importance of autonomy from the
student perspective. In order to counteract the
impression that DSA could limit the autonomy of
students, there is quite a bit that can be done on the
developer's side. With regard to wording, it is
advisable to ensure that proposals are not made in a
patronizing or commanding tone. Also, too intrusive
reminders and categorization of students should be
avoided in order not to create reactance among users.
Ethical Perception of a Digital Study Assistant: Student Survey of Ethical Values
101
Ideally, students will see the DSA as a helpful tool,
which is proactive, but still discrete, respectful and
accepts personal decisions. It is not surprising that the
categories transparency/informed consent, privacy,
and data security are highly correlated with each
other, as already stated in the category data
privacy/anonymity. The correlations indicate a strong
connection between these categories. Simplified it
can be said that students want to know what happens
to their data, expect that the declared purpose of the
data use will be adhered to, and attach great
importance to the protection of their data from theft
or third-party access.
6 CONCLUSION & FUTURE
WORK
In this article, two studies were combined in order to
examine important ethical values perceived by
students in the context of a DSA. For this purpose,
five ethical value categories were first derived by
study 1 via free text answers. Afterwards 156 students
were surveyed with regard to these categories in study
2. This paper is intended to provide initial indications
of which ethical values are particularly important to
students when using a DSA and what should be taken
into account when developing such a system.
This work can be understood as a first step
towards incorporating concepts of ethical values or
VSD into the development of digital assistance
systems for students. It is not intended to claim
completeness of the ethical values, nor does this
research explicitly search for reasons or possible
implementation methods. This opens up interesting
perspectives for further research in the field of higher
education in general and research on digital study
assistants in particular. A next logical step would be
to investigate the implementation of ethical values in
a DSA. In other words, how does the system manage
to address and consider the ethical values of students?
Furthermore, follow-up research with students who
are actual using the DSA in their daily study routine
would be interesting and would offer further
insightful implications for researchers and
practitioners. Developers and decision-makers can
use this paper as a basis for their decision to include
ethical considerations in the development of systems
that are used by students and to take their ethical
values into account.
REFERENCES
Abbasi, S., Kazi, H.(2014). Measuring Effectiveness of
Learning Chatbot Systems on Student’s Learning
Outcome and Memory Retention. Asian Journal of
Applied Science and Engineering 3(2), 251–260.
Ahel, O. and Lingenau, K. (2020). Opportunities and
Challenges of Digitalization to Improve Access to
Education for Sustainable Development in Higher
Education. In: W. Leal Filho, A. L. Salvia, R. W.
Pretorius, L. L. Brandli, E. Manolas, F. Alves, A. Do
Paco (eds.) Universities as Living Labs for Sustainable
Development: Supporting the Implementation of the
Sustainable Development Goals, 341–356, Cham:
Springer International Publishing.
Alexander, B., Ashford-Rowe, K., Barajas-Murphy, N.,
Dobbin, G., Knott, J., McCormack, M., Pomerantz, J.
and Seilhamer, R. (2019). Educause Horizon report:
2019 Higher Education edition.
Allemann-Ghionda, C. (2014). Internationalisierung und
Diversität in der Hochschule. Zeitschrift für
Pädagogik, 5, 668–680.
Anderson, M. and Anderson, S.L. (2011). Machine Ethics.
New York: Cambridge University Press.
Apt, W., Schubert, M. and Wischmann, S. (2018). Digitale
Assistenzsysteme -Perspektiven und
Herausforderungen für den Einsatz in Industrie und
Dienstleistungen, Institut für Innovation und Technik
(iit) Berlin, 12-42.
Benotti, L., Martínez, M., Schapachnik, F (2014). Engaging
High School Students using Chatbots. In: Proceedings
of the 2014 Conference on Innovation & technology in
Computer Science Education, 63-68. Association for
Computing Machinery, Uppsala.
Brandenburg, S., Minge, M., Cymek, D. (2017). Common
Challenges in Ethical Practice when Testing
Technology with Human Participants: Analyzing the
Experiences of a Local Ethics Committee. i-com 16(3),
267-273.
Brandenburg, S., Schott, R., Minge, M. (2018). Zur
Erfassung der ethischen Position in der
Softwareentwicklung. In: Dachselt, R., Weber, G. (eds.)
Mensch und Computer 2018 Workshopband, 265-
273. Gesellschaft für Informatik e.V., Bonn.
British Academy and Royal Society (2018). Data
Management and Use: Governance in the 21st Century.
Cohen, J. E. (2000). Examined lives: Informational Privacy
and the Subject as Object. Stanford Law Review 52(5),
1373-1438.
Drachsler, H., Hoel, T., Scheffel, M., Kismihok, G., Berg,
A., Ferguson, R., Manderveld, J. (2015). Ethical and
Privacy Issues in the Application of Learning Analytics.
5th International Learning Analytics & Knowledge
Conference (LAK15): Scaling Up: Big Data to Big
Impact, 390-391. Association for Computing
Machinery, New York.
Dutta, D. (2017). Developing an Intelligent Chatbot Tool to
assist High School Students for Learning General
Knowledge Subjects. Georgia Institute of Technology,
Atlanta.
CSEDU 2022 - 14th International Conference on Computer Supported Education
102
Fitzgerald, M., Kruschwitz, N., Bonnet, D. and Welch, M.
(2013). Embracing digital technology: A new strategic
imperative. MIT Sloan Management Review, 55, 1–12.
Flanagan, M., Howe, D. C., Nissenbaum, H. (2008).
Embodying Values in Technology: Theory and
Practice. In: van den Hoven, MJ., Weckert, J. (eds.)
Information Technology and Moral Philosophy, 322-
353. Cambridge University Press, New York.
Floridi L (2010). Ethics after the information revolution. In:
Floridi L (Ed.) The Cambridge handbook of
information and computer ethics. Cambridge:
Cambridge University Press.
Floridi, L. (2015). The Ethics of Information. Oxford:
Oxford University Press.
Friedman, B., Kahn, P., Borning, A. (2013). Value Sensitive
Design and Information Systems. In: Doorn N.,
Schuurbiers D., van de Poel I., Gorman M. (eds.) Early
Engagement and New Technologies: Opening up the
Laboratory. Philosophy of Engineering and
Technology 16, 348-372. Springer, Dordrecht.
Gasser, M. (1988). Building a Secure Computer System.
Van Nostrand Reinhold, New York.
Gottburgsen, A. and Wilige, J. (2018). Mehr
Mobilitätserfahrungen durch digitale Medien? Zu den
Effekten von studentischer Diversität und Lernumwelts-
merkmalen auf die internationale Mobilität. Beiträge
zur Hochschulforschung, 4/2018, 30–50.
Harris, I., Jennings, R., Pullinger, D., Rogerson, S.,
Duquenoy, P. (2011). Ethical assessment of new
technologies. A meta-methodology. Journal of
Information Communication and Ethics in Society 9(1),
49–64.
Heuchemer, S. and Sander, H. (2018). Hochschulbildung
4.0 – (K)eine Abkehr von bisherigen Bildungszielen?!
In: Auferkorte-Michaelis N., Linde F. (eds.) Diversität
lernen und lehren ein Hochschulbuch, Opladen,
Berlin, Toronto: Verlag Barbara Budrich.
Hirsch-Kreinsen, H., Ittermann, P. and Niehaus, J. (2018).
Digitalisierung industrieller Arbeit: die Version
Industrie 4.0 und ihre sozialen Herausforderungen. 2st
edn. Baden-Baden Nomos.
Hold, P., Erol, S., Reisinger, G. and Sihn, W. (2017).
Planning and Evaluation of Digital Assistance Systems.
Procedia Manufacturing, Bd. 9, 143-150.
Ienca, M., Wangmo, T., Jotterand, F., Kressig, R., Elger, B.
(2017). Ethical Design of Intelligent Assistice
Technologies for Dementia: A Descriptive Review. Sci
Eng Ethics 24(4), 1035-1055.
Ihme, J. M. and Senkbeil, M. (2017). Warum können
Jugendliche ihre eigenen computerbezogenen
Kompetenzen nicht realistisch einschätzen?, Zeitschrift
für Entwicklungspsychologie und Pädagogische
Psychologie 49(2), 24–37.
Johann, T., Maalej, W. (2013). Position Paper: The Social
Dimension of Sustainability in Requirements
Engineering. Engineering. In: In Proceedings of the 2nd
International Workshop on Requirements Engineering
for Sustainable Systems, 1-3.
Johnson, L., Adams Becker, S., Estrada, V. and Freeman,
A. (2014). NMC Horizon Report: 2014 Higher
Education Edition. Austin: The New Media
Consortium.
Johnson, D. G. and Miller, K. W. (2009). Computer Ethics:
Analyzing Information Technology. Pearson Education
International, Upper Saddle River.
Keber, T., Bachmeier, E., Neef, K. (2019). Learning
Analytics – Datenschutzrechtliche und ethische
Überlegungen zu studienleistungsbezogenen Daten-
analysen an Hochschulen. JurPC Web-Dok 97, 1-72.
Kernaghan, K. (2014). Digital Dilemmas: Values, Ethics
and Information Technology. Canadian Public
Administration 57(2), 295-317.
Leal Filho, W., Salvia, A. L., Pretorius, R. W., Brandli, L.
L., Manolas, E., Alves, F., Azeiteiro, U., Rogers, J.
(2020). Universities as Living Labs for Sustainable
Development: Supporting the Implementation of the
Sustainable Development Goals. Springer Nature
Switzerland AG
Lin, P., Abney, K. (2017). Robot Ethics 2.0: From
Autonomous Cars to Artificial Intelligence. New York:
Oxford University Press.
Manzeschke, A., Brink, A. (2020). Ethik der
Digitalisierung in der Industrie. In: Frenz, W. (eds.)
Handbuch Industrie 4.0: Recht, Technik, Gesellschaft,
1383–1405. Springer, Berlin, Heidelberg.
Masrom, M., Ismail, Z., Hussein, R., Mohamed, N. (2010).
An Ethical Assessment of Computer Ethics Using
Scenario Approach. International Journal of Electronic
Commerce Studies 1(1), 25-36.
Mayring, P. (2015). Qualitative Inhaltsanalyse:
Grundlagen und Techniken, 12, Beltz.
Millar, J. (2016). An Ethics Evaluation Tool for Automating
Ethical Decision-Making in Robots and Self-Driving
Cars. Applied Artificial Intelligence 30(8), 787-809.
Moor, J.H. (1985). What is Computer Ethics?.
Metaphilosophy 16(4), 266–275.
Morakanyane, R., Grace, A. and O’Reilly, P. (2017).
Conceptualizing digital transformation in business
organizations: A systematic review of literature.
Proceedings of the 30th Bled eConference, 427–443.
Mukerjee, S. (2014). Agility: A crucial capability for
universities in times of disruptive change and
innovation. Australian Universities’ Review, 56, 5.
Munoko, I., Brown‑Liburd, H. L., Vasarhelyi, M. (2020).
The Ethical Implications of Using Artificial Intelligence
in Auditing. Journal of Business Ethics 167, 209-234.
Pardon, A. and Siemens, G. (2014). Ethical and Privacy
Principles for Learning Analytics. British Journal of
Educational Technology 45(3), 438-450.
Ropohl, G. (1996). Ethik und Technikbewertung. 2nd edn.
Suhrkamp Verlag, Frankfurt am Main.
Rouse, W. B. (2005). A Theory of Enterprise
Transformation. Systems Engineering 8(4), 279–295.
Schmohl, T. and Schäffer, D. (2019). Lehrexperimente der
Hochschulbildung: didaktische Innovationen aus den
Fachdisziplinen.
wbv Media GmbH & Co. KG,
Bielefeld.
Shum, S. B., Ferguson, R. (2012). Social Learning
Analytics. Journal of Educational Technology and
Society 15(3), 3–26.
Ethical Perception of a Digital Study Assistant: Student Survey of Ethical Values
103
Slade, S., Prinsloo, P.(2013). Learning Analytics: Ethical
Issues and Dilemmas. American Behavioral Scientist
57(10), 1510–1529.
Steinmann M., Shuster, J., Collmann, J. (2015). Embedding
Privacy and Ethical Values in Big Data Technology. In:
Transparency in Social Media, 277–301, Springer,
Switzerland.
van den Hoven, J. (2017). Ethics for the Digital Age: Where
Are the Moral Specs?. In: Werthner H., van Harmelen
F. (eds.) Informatics in the Future: Proceedings of the
11th European Computer Science Summit (ECSS
2015), 65-76. Springer, Cham.
Willis, J. E.(2014). Learning Analytics and Ethics: A
Framework Beyond Utilitarianism. Educause Review.
Wright, D. (2011). A Framework for the Ethical Impact
Assessment of Information Technology. Ethics and
Information Technology 13(3), 199–226.
Zervakis, P. and Mooraj, M. (2014). Der Umgang mit
studentischer Heterogenität in Studium und Lehre.
Chancen, Herausforderungen, Strategien und
gelungene Praxisansätze aus den Hochschulen. In:
Zeitschrift für Inklusion, (1-2).
CSEDU 2022 - 14th International Conference on Computer Supported Education
104