The Future of Higher Education Is Social and Personalized!
Experience Report and Perspectives
Sven Strickroth
and Franc¸ois Bry
Institute for Informatics, Ludwig-Maximilian University of Munich, Germany
Technology-enhanced Learning, Peer Review, Automatic Feedback, Student Activation, Learning Analytics,
This position paper is devoted to learning and teaching technologies aimed at alleviating problems and ad-
dressing challenges faced by learners or teachers in mass classes. It reports on practical experiments run
over the last terms at the Institute for Informatics of Ludwig-Maximilian University of Munich, Germany.
Furthermore, perspectives for higher education opened by technology are discussed: Digital social learning
and/or experimentation spaces and technology-based personalization. It is further argued that such approaches
provide specific advantages that are not only desirable for teaching mass classes.
Higher education in most countries of continental
Europe and other world regions is nowadays char-
acterized by mass classes, that is, courses attended
by one to several hundred students. According to
the constructivism learning theory and connectivism
(Goldie, 2016), learning is an inherently social pro-
cess. However, social interactions and discussions be-
come rather limited when it comes to mass teaching.
Mass classes make frontal lectures a last resort what
often results in rather inactive students and high drop-
out rates. Furthermore, mass classes often limit the
feedback that can be provided by teachers and tutors
and make the grading of examinations lengthy and
therefore challenging. This position paper is devoted
to learning and teaching technologies aimed at alle-
viating the aforementioned problems and addressing
the outlined challenge of mass teaching.
This position paper reports on experiences gained
at the Institute for Informatics of Ludwig-Maximilian
University of Munich, Germany, through deploy-
ing both established and novel Technology-Enhanced
Learning (TEL) methods to alleviate many disad-
vantages of mass teaching for students and teachers
alike in three dimensions: First, approaches to acti-
vate students in mass classes such as learning-specific
backchannels and peer teaching; second, approaches
to crowdsource teaching tasks such as giving feed-
back, correction of submitted solutions and support-
ing tutors for collaborative feedback provisioning;
third, automatized feedback provisioning and exam-
ination pre-corrections.
Furthermore, this article reports on experiences
of applying data science in education investigated at
the same institute: Improving learners’ self-regulated
learning by nudging them to peer reviews, by report-
ing on their learning activities, and by predicting the
correlations of their examination performances with
their learning activities and, finally, detection of sys-
temic errors among learners by human computation
and collaboration.
The research questions addressed in this paper are:
(1) How to activate students and increase interactitity
in mass classes? (2) How can large numbers of stu-
dents be exploited (e.g., croudsource teaching tasks)?
(3) How to (semi)-automate teaching tasks such as
feedback provisioning and correcting? and (4) How
to nudge students to active learning?
Finally, perspectives for exploiting TEL in higher
education are presented. The discussed perspectives
are: The need for social learning spaces, personaliza-
tion and spaces for experimentation and discussions.
The contributions of this position paper are an
overview of deployed approaches to tackle the afore-
mentioned challenges and perspectives for TEL-
based social and personalized learning in higher ed-
Strickroth, S. and Bry, F.
The Future of Higher Education Is Social and Personalized! Experience Report and Perspectives.
DOI: 10.5220/0011087700003182
In Proceedings of the 14th International Conference on Computer Supported Education (CSEDU 2022) - Volume 1, pages 389-396
ISBN: 978-989-758-562-3; ISSN: 2184-5026
2022 by SCITEPRESS Science and Technology Publications, Lda. All rights reserved
One of the main issues with mass lectures is the
limited interactivity. In this section, experiences for
two approaches are outlined: (1) learning-specific
backchannels from the students to the teacher and (2)
TEL approaches to activate students during lectures.
One of the goals for developing the tool Back-
stage (Gehlen-Baum et al., 2012) was to provide a
connected and interacting community dedicated to a
single course. The reason for not using existing so-
cial media, such as Twitter, was to avoid distraction
from off-topic content and to exploit novel learning-
specific features. The first version of Backstage of-
fered a digital backchannel to connect lectures’ par-
ticipants (both students and teachers alike). Back-
stage provided a forum-like structure with which stu-
dents could ask questions, answer questions of their
peers, and comment on the lecture or annotate lec-
ture slides. Posts by students could be up and down
voted by the students and teachers as well as catego-
rized by the lecture participants as ”question”, ”an-
swer”, ”remark”, and ”off-topic”. This tagging has
shown to be effective at preventing an off-topic use
of the backchannel through a social control by the
student community itself. Backstage has proven to
foster interactivity and awareness in large-class lec-
tures and to encourage lecture-relevant communica-
tion in different courses (Bry and Pohl, 2017). Com-
pared to a simple chat such as provided by video-
conference systems like Zoom Backstage’s backchan-
nel is much more effective at focusing the communi-
cation at lecture-relevant contents and at creating so-
cial interactions like several students progressively re-
fining a question to a teacher or answering questions
posed by fellow students.
Figure 1: Domain-specific puzzle-like question; taken from
(Mader and Bry, 2019a), Fig. 4, p. 213.
A second version of Backstage (Mader and
Bry, 2019a) has extended the system with an au-
dience response system (ARS) supporting sophisti-
cated quizzes. Backstage’s ARS is aimed at acti-
vating students within and without lectures. Back-
stage is a web-based system and can therefore be used
by students with their smartphones or mobile com-
puters. Unlike simple ARS, various quiz types are
supported, ranging from simple multiple-choice ques-
tions over the identification of hot-spots in images
(useful for, e.g. medicine education or geography) to
domain-specific question types such as creating logic
proofs and puzzle-like programming tasks (cf. Fig-
ure 1). Furthermore, Backstage’s quizzes are adaptive
regarding the student’s needs and knowledge. Back-
stage also supports ”phased quizzes” which can run
over a longer period of time so as to support self-
learning. On the one hand Backstage’s quizzes give
teachers an immediate feedback on the students’ un-
derstanding and competencies. On the other hand,
they also activate students and give them instant feed-
back. Backstage, like other ARS in general, have
been proven to have positive effects on attendance
and engagement (Kay and LeSage, 2009; Oigara and
Keengwe, 2013) and can also be used to initiate and
foster (offline) peer discussion (Mazur and Somers,
1999) in class.
The main issue with mass teaching is that the num-
ber of students is increasing but the number of teach-
ers and tutors cannot be increased accordingly due to
limited resources. This directly limits the provision-
ing of (timely) feedback, which, according to Hattie
is one of the most important factors for learning suc-
cess (Hattie and Timperley, 2007). However, a large
class also makes it possible to leverage manpower by
crowdsourcing certain teaching tasks.
A first approach to cope with an insufficient num-
ber of teachers is peer review. Peer review is a learn-
ing activity in which students evaluate and deliver
written feedback on the work of their peer students
(Nicol, 2010). Peer review and peer feedback is of-
ten used in essay writing (Cho and Schunn, 2007),
however, it can be used if the tasks to be reviewed
require some creativity. Peer review can be operated
asynchronously on and as homework, synchronously
in class, as a pre-correction for the teacher, or as
the final feedback for the students. Peer review and
peer feedback have been successfully tested with the
systems Backstage (Heller and Bry, 2019b; Mader
and Bry, 2019d) and GATE (Strickroth et al., 2011),
designed and deployed by the authors in different
scenarios such as introductory programming courses
(teaching Haskell and Java). The authors’ experience
with programming assignments is that students good
at spotting their peers’ errors are also good at iden-
tifying correct peers’ submissions and therefore are
little prone to give false feedback (Heller and Bry,
CSEDU 2022 - 14th International Conference on Computer Supported Education
2019b). However, about 22 % of the given feedback
turned out to be partially incorrect further research
on how to support students is necessary. In general,
peer review has shown to be a powerful method as it
allows students to receive timely and extensive feed-
back; furthermore, multiple reviews by different peers
have been shown to be better than a single review of
a expert (Cho and Schunn, 2007). Peer feedback and
review also have other advantages that are discussed
in Section 5.
Another approach to providing feedback has been
tested with a prototype system for reducing the teach-
ers’ and tutors’ workload for correcting homework
(Heller and Bry, 2021). This approach relies on the
fact that (1) mass teaching also requires multiple tu-
tors who correct submitted solutions and (2) students
often make similar mistakes (e. g., common miscon-
ceptions). In order to support teachers, the proposed
software uses text processing, collaborative filtering,
and teacher collaboration in a wiki-like environment
and suggests feedback fragments directly while typ-
ing them in. The approach has demonstrated its fea-
sibility and its effectiveness in multiple case studies
(Heller and Bry, 2021).
Learning requires assessment (Kirkwood and Price,
2008). One can distinguish between assessment for
learning (also often referred to as ”formative assess-
ment”) and assessment of learning (also often re-
ferred to as ”summative assessment”). In this sec-
tion experiences with two formative assessment ap-
proaches (automated feedback generation and in-
class feedback for the teacher) and one summative
assessment approach using computer-assisted pre-
corrections which both have been tested in two pro-
gramming courses (in Java and Haskell, respectively)
are reported.
Learning requires that students do their homework
regardless of whether the homework is correct or in-
correct. In addition, learning is improved by immedi-
ate and delayed feedback (Narciss, 2008). Hence, an
approach combining both types of feedback has been
It is worth stressing that both immediate and de-
layed feedback are hardly possible in large classes
without the help of technology. For programming
tasks, the approach tested consisted of two automated
checks that could be requested twice by every student
before the final submission deadline. This limitation
to two requests aimed at preventing misuses such as
gaming-the-system approaches (Baker et al., 2008).
The first check was a syntax check using a com-
piler. The second check was a unit test. A former
study with the GATE system has shown that even
simple feedback such as knowledge of result/response
(Narciss, 2008) (i.e., whether a program compiles or
a unit test passed or failed) where each test can be re-
quested by students only once significantly reduced
errors in their final submissions (Strickroth et al.,
2011). Additionally, the test results were presented
to the tutors for each submission as an aid for their
manual inspection. This was rated as very helpful by
the tutors. In both courses, the correct solutions were
discussed in dedicated sessions.
Such automated evaluations can also be used in
synchronous settings: A challenge in large classes is
the identification of students who need help the most.
This is particularly the case in synchronous program-
ming labs: It turns out that it is not sufficient to have
a teacher and a few tutors going through the seating
rows to help students. A majority of students has
questions and expects feedback. Therefore, it often
happens that teachers do not recognize, or do not rec-
ognize early enough, the most struggling students. To
address this issue, a learning analytics dashboard was
developed for Backstage that provides teachers with
a real-time overview of the progress of all students
(Mader and Bry, 2019c). This dashboard allowed the
most struggling students to be easily identified by the
teacher. The results of a comparative study in which
the progress of all students with and without this dash-
board were available, are promising (Mader and Bry,
Finally, this sections reports on recent experi-
ments to support mass examinations for two differ-
ent programming courses (with Haskell and Java, re-
spectively). Corrections and gradings should be com-
pleted within a specific time frame so that the students
can take notice of their performances early enough
and, if necessary, can register and learn for a second-
chance examination. Traditional paper-and-pen ex-
aminations have several drawbacks: First, handwrit-
ten text is rarely easy to read. Second, open-ended
questions are complicated to grade, and are also prone
to lengthy discussions during the inspection of the
graded examination in Germany, students have the
right to inspect their graded examination and to raise
any objections. Therefore, a fully digital decentral-
ized open-book examination using the GATE system
was used. The questions were designed as fill-in-the-
gap tasks with one to a few correct solutions or as
multiple-choice questions. This also had the side-
The Future of Higher Education Is Social and Personalized! Experience Report and Perspectives
effect that the examination was more realistic and
competency-oriented as the students do not need to
program on a sheet of paper with a pen, but can use
a compiler and test their solutions. Another design
choice was to use a binary marking schema without
issuing partial marks. This way the submitted solu-
tions could be evaluated automatically in real-time
this also provided the teachers a real-time view on the
progress of the students during the examination. A fi-
nal human verification of every answer marked by the
software as ”probably incorrect” ensured a fair and
thorough correction. The automatic pre-correction
ensures that only a fraction of the submitted solutions
needed to be inspected by the tutors which resulted in
a significant reduction of the time needed for the cor-
rection. In a first investigation, no significant differ-
ence to traditional examinations regarding the pass-
failure-rates or achieved grades could be observed.
In this section three approaches to nudging students
to actively engage in learning are presented. These
go beyond common approaches to nudge students to
their homework such as is declaring the homework a
condition of participation in an examination or to by
grading based on both homework and examination.
Both traditional approaches are not always possible
due to examination regulations.
Peer review can not only be used to crowdsource
teacher tasks (cf. Section 3) but also to provide feed-
back which improves the learning process by acti-
vating students (Li et al., 2019; Zheng et al., 2019).
However, an important prerequisite for using peer re-
view is that students actually write feedback and not
just try to get feedback from others. A low partici-
pation rate poses a significant threat for the entire ap-
proach, as students may receive little to no feedback.
A first experiment tested with Backstage in a
Bachelor’s introductory programming course a low
participation rate was observed, however, most stu-
dents indicated that receiving and delivering peer re-
views was “mostly helpful” to their learning and that
seeing and thinking about different solutions was re-
marked positively by the students (Heller and Bry,
In a second experiment currently underway using
GATE, a restriction was introduced to motivate stu-
dents to submit their solutions and to deliver feedback
regularly (skipping twice results exclusion from fur-
ther peer feedback). Students’ responses to peer feed-
back are promising: Participation in peer reviewed as-
signments is higher than participation to assignments
which are not peer reviewed. This suggests that re-
viewing different peers’ solutions and receiving feed-
back from peers probably is a motivating factor to do
the assignments. An in-depth analysis of the collected
data is underway.
Audience response systems also can not only im-
prove interactivity in class but can also nudge students
to actively participate in a lecture and reflect on spe-
cific problems. Studies suggest that the use of ARS
can lead to higher retention rates and also has pos-
itive effects on overall class examination scores, es-
pecially for students whose performances are in the
lower quartile (Hoyt et al., 2010; Kay and LeSage,
2009). Using Backstage, a team-based social com-
petition with quizzes aimed at boosting participation
were tested (Mader and Bry, 2019b). A first evalua-
tion in a small class demonstrated the effectiveness of
the approach, and a second evaluation suggests that
for use in large classes teams have to be built in a spe-
cific way (Mader and Bry, 2019b).
Finally, this sections reports on a case study that
uses learning analytics to encourage students not to
skip homework and to increase their participation in
class (Heller and Bry, 2019a). In a Bachelor course
on theoretical computer science, students were given
individual predictions of their withdrawal, or “skip-
ping” of assignments, and their examination perfor-
mances (called ”examination fitness”). The evalua-
tion shows that this course had the lowest skipping
rates compared with the same courses over the for-
mer three years for two years the difference is also
statistically significant (Heller and Bry, 2019a). Atti-
tudes toward such predictions were also investigated:
The students did not find any of the predictions dis-
couraging, nor did they report being motivated to a
higher participation to do homework. However, the
examination fitness prediction was perceived as more
interesting than the skipping prediction.
While TEL is often used to address only pressing
classroom organization and management issues (e.g.,
managing examination registrations and distributing
PDF files) (Markova et al., 2017; Henderson et al.,
2017), the experience reported in the previous section
shows that TEL can also be used to support social as-
pects of learning, such as computer-supported collab-
orative learning, co-regulation and peer teaching. In
this section the authors present their vision for TEL-
based social and personalized learning in higher edu-
cation that is not only relevant to mass teaching.
CSEDU 2022 - 14th International Conference on Computer Supported Education
6.1 Social Learning Spaces
Good learning, and therefore good teaching, require
an intensive, active exchange among students as well
as between students and teachers. Such an exchange,
however, is what disappears first in mass teaching e. g.
when there are hundreds of students in a course. Not
all students dare to raise issues or ask questions in
front of large audiences. Moreover, it can also be dif-
ficult to find learning partners when you sit next to a
different student in every lecture.
Social media specialized for learning can be used
to create social learning spaces that would be oth-
erwise hardly possible. As described in the previ-
ous sections, such systems allow students to ask and,
more importantly, structure questions and possible an-
swers. Structuring can also be improved by allow-
ing students to vote for specific questions which they
share. This way, the teacher does not have to deal with
a dozen questions, but can gain a quick overview and
focus on the most urgent ones. Similar approaches
could also be used in asynchronous scenarios where
students have a shared digital workspace like a wiki
where they can collaboratively collect and optimize
their questions and answers or provide peer feedback.
TEL allows to mitigate the bottleneck in communica-
The last example shows that it is possible to turn
mass teaching into an advantage: A large number of
students makes it possible to involve them and take
advantage of their heterogeneity by forming a com-
munity of practice (Wenger et al., 2002). The students
have a common goal and can work collaboratively to
solve difficult problems. The greater heterogeneity in
the group of students is also likely to lead to more
questions, and again, collaboration can lead to better
questions and to standing up for each other (among
others, peer reviews or collaboratively collecting, de-
veloping, and optimizing learning materials e. g. in
a wiki). Technology also allows students at different
locations to collaborate who are not at the same uni-
versity or even not on the same continent.
The challenge is to design that community and to
foster the exchanges among students and between
students and teachers alike both on the pedagogi-
cal and technical sides. Furthermore, there are many
more social aspects that can be supported such as how
to find other students who share the same interests or
are struggling with the same issues.
There are two further issues to address: First,
given the speed at which STEM (Science, Technol-
ogy, Engineering, and Mathematics) is evolving, life-
long learning is becoming increasingly important.
Therefore, universities should go beyond alumni por-
tals where graduates have to register manually and
create social spaces that allow active exchange with
their (former) students to learn and collaborate.
Second, social (learning and collaboration) spaces
for teachers are also needed that go beyond a single
department, university, or even country. Professional
social media can enable teachers with similar interests
to connect and share experience as well as collabo-
ratively develop teaching materials (Bothmann et al.,
2021; Strickroth et al., 2015). For example, special-
ized systems and repositories that allow peer review-
ing and collaboratively optimizing examination ques-
tions should be considered. Such approaches are not
possible without the use of technology.
6.2 Personalization
Personalization is probably best achieved by teachers
who are in direct contact with their students (cf. pre-
vious section). However, such direct interaction is not
always possible, e. g., with a students-to-teacher ratio
over 800 for professors, and over 70 for teaching as-
sistants (Heller and Bry, 2019b). Here, automation
and support through technology can not only free up
time that can then be used for more interaction with
students, but also provide new insights (on both the
small and larger scale) that would hardly be possible
without technology.
For learning, students build their own digital per-
sonal learning environment with systems provided by
the university, such as learning management systems
or systems described in the previous sections, and
with systems that they selected for specific tasks. Ev-
ery interaction with these systems such as solving
(manual or auto generated) assignments, data traces
are generated on the underlying systems. Not only
do students generate data when they interact with the
systems, but so do teachers when they assess student
contributions. This data is often not used systemati-
cally, and if it is, it is only used within a single system.
The data collected can be useful in two ways:
First, students’ submitted solutions, previous at-
tempts and attached meta-data such as marks, grades,
or feedback given by teachers can be used for person-
alization on a small scale: Learning analytics can be
applied to the available data in a system to adapt to the
needs and knowledge levels of specific students or to
aid tutors to give timely feedback. There is a wide
spectrum for possible personalization approaches, in-
cluding generating feedback (e.g., based on similar
submissions or shared misconceptions in real-time),
generating questions based on the detected miscon-
ceptions or knowledge gaps, and using the data of
previous venues as a basis for predictions in subse-
The Future of Higher Education Is Social and Personalized! Experience Report and Perspectives
quent venues of courses to warn and support students
at risk. Pre-corrections and learning analytics can also
be used to support tutors or their collaboration while
inspecting student solutions. Additionally, one can
analyze how the students interact with the system and
with each other, identify usage and learning patterns
that can help to adjust the teaching method and to
optimize the learning environment including the used
Second, students’ (life cycle) data such as their
participation in exams, their grades in previously at-
tended courses, etc. can be harnessed using data sci-
ence techniques. On the one hand, this makes it possi-
ble to identify students at risk, providing personalized
warnings and recommendations for their studies. On
the other hand, it can also be used to identify difficult
courses and better prepare students for these.
Besides advantages, there are also challenges:
Such Learning analytics approaches have only been
rarely used in Europe, possibly because of a
widespread fear among European higher education
teachers to violate privacy regulations. Local uni-
versity or statewide initiatives are needed to establish
trusted learning analytics that engage all stakehold-
ers (cf. (Drachsler and Greller, 2016)). A good start-
ing point could be one study program and then, ex-
tend it other study programs and then develop plans
on how the data can be exchanged across institutions
(e.g., school to university). In addition to privacy, so-
lutions are also needed to avoid possible bias in the
data (Riazy and Simbeck, 2019) and to train teachers
and students on how to interpret and make use of the
analyses (Slade and Prinsloo, 2013).
6.3 Space for Experimenting and
Finally, spaces for experimenting with innovative ap-
proaches to teaching are desirable. This needs to be
seen on two different levels:
First, a change in the culture (at least in continen-
tal Europe) regarding teaching is needed. Teaching is
often not discussed between teachers outside the al-
ready interested communities. There are also inhibi-
tions to experimentation, such as data protection reg-
ulations that are perceived as too complicated. The
COVID-19 pandemic triggered many experiments,
however, teachers should not completely fall back to
old habits. Rather, teachers should learn from their
attempts to deal with teaching in the pandemic and
use this as a foundation for further discussion and
experimentation. There is an enormous pedagogical
potential. Teachers also need to be more creative to
think of scenarios that are not or hardly possible with-
out the use of technology such as making use of the
heterogeneity of a large class. Again, an exchange
between teachers and Technology-Enhanced Learn-
ing researchers is necessary to join forces to build up
and develop new ideas (cf. Section 6.1). A side ef-
fect, or even a specific goal, could be that the tech-
nology enables people with disabilities or otherwise
time/location-constrained persons such as single par-
ents to take part in courses or learning scenarios.
Second, (flexible) technological infrastructures
are needed! This is a rather technical point of view
but it is equally important as the first point. On the
one hand an infrastructure is required so that also non-
tech-savvy teachers can set up and use software that
fits their needs in a privacy-conform manner that runs
within the university (Strickroth et al., 2021). On the
other hand there will be prototypes that evolve over
time and show to be effective. A challenge here is to
bring these into production and transfer the operating
to data centers as researchers cannot operate technol-
ogy for a whole university (Bußler et al., 2021; Kiy
et al., 2017).
In practice TEL is often used to address only press-
ing classroom organization and management issues,
however, the experience reported in this position pa-
per shows that TEL can not only be used to support
social aspects of learning but can also turn certain as-
pects of mass teaching into an advantage.
The paper discussed experiences and different ap-
proaches for enabling interaction in mass classes,
crowdsourcing teaching tasks, automated feedback
and computer-assisted correction, and nudging stu-
dents to active learning. Finally, the paper presents
perspectives on digital social learning and/or experi-
mentation spaces and technology-based personaliza-
tion for higher education opened by technology. The
experiments presented in this paper are first steps to-
wards this vision.
The presented results also outline further direc-
tions for researching and optimizing the described ap-
proaches such as how to better support students in
peer reviews on programming assignments. The em-
ployed technologies are research prototypes. Limita-
tions and drawbacks are discussed in the respective
papers and cannot be presented in detail here due to
page limitations.
Almost all studies deal with Computer Science
contexts and, therefore, employ domain specific ap-
proaches such as unit tests for automatically evaluat-
CSEDU 2022 - 14th International Conference on Computer Supported Education
ing programming assignments that are not available
in other domains. However, the authors argue that
comparable automatic tests can be used in STEM con-
texts. For other domains such as languages heuris-
tics based on metrics or upcoming machine learning
approaches might be applicable as an aid for semi-
automated grading or formative feedback (e.g., (Stab
and Gurevych, 2017)). Nevertheless, such approaches
need to be further investigated. ARS and peer re-
view have already been used in various contexts (e.g.,
(Keough, 2012; van Popta et al., 2017)).
Most of the approaches described are not lim-
ited to mass teaching but can also be used in smaller
classes. Here, related disciplines or practices in a
school context with about 30 learners are classroom
orchestration (cf. (Dillenbourg, 2013)) and learning
engineering (cf. (Baker et al., 2021)) which can-
not be discussed in detail here due to page limita-
tions. Note, however, that the use of technology
should not replace the human component in learn-
ing and teaching but should enable or comprehend
it. That means that technology can provide timely
personalized feedback to students, allow students at
different locations to collaborate, enable disabled or
time/locations-constrained students to learn and inter-
act with each other, free up time from certain (often
recurring and/or tedious) tasks by using automation,
or to provide insights into learning processes etc. that
are hardly possible otherwise. TEL offers an enor-
mous social and pedagogical potential that needs to
be explored. . .
The authors are thankful to over 10.000 students who,
over the last decade actively participated in testing
and discussing the approaches to learning and teach-
ing reported about in this article. The authors are
also thankful to their colleagues and (doctoral) stu-
dents, especially to Dr. Niels Heller and Dr. Sebastian
Mader, for their feedback, advice, and contributions
to the research reported about in this article. The au-
thors also thank the reviewers for their valuable feed-
Baker, R., Walonoski, J., Heffernan, N., Roll, I., Corbett,
A., and Koedinger, K. (2008). Why students engage
in “gaming the system” behavior in interactive learn-
ing environments. Journal of Interactive Learning Re-
search, 19(2):185–224.
Baker, R. S., Boser, U., and Snow, E. L. (2021). Learning
engineering: A view on where the field is at, where
it’s going, and the research needed. Available online:
TMB Learning%20Engineering v12rsb.pdf.
Bothmann, L., Strickroth, S., Casalicchio, G., R
ugamer, D.,
Lindauer, M., Scheipl, F., and Bischl, B. (2021). De-
veloping open source educational resources for ma-
chine learning and data science. arXiv:2107.14330v2.
Bry, F. and Pohl, A. Y.-S. (2017). Large class teaching with
Backstage. Journal of Applied Research in Higher Ed-
ucation (Special Issue on reviewing the performance
and impact of social media tools in higher education),
Bußler, D., Lucke, U., Strickroth, S., and Weihmann, L.
(2021). Managing the transition of educational tech-
nology from a research project to productive use.
In SE-SE2021 - Software Engineering 2021 Satel-
lite Events - Workshops and Tools & Demos. Pro-
ceedings of the Software Engineering 2021 Satellite
Events., volume 2814 of CEUR Workshop Proceed-
Cho, K. and Schunn, C. D. (2007). Scaffolded writing
and rewriting in the discipline: A web-based recip-
rocal peer review system. Computers & Education,
Dillenbourg, P. (2013). Design for classroom orchestration.
Computers & Education, 69:485–492.
Drachsler, H. and Greller, W. (2016). Privacy and analytics:
it’s a DELICATE issue a checklist for trusted learning
analytics. In Proceedings of the sixth international
conference on learning analytics & knowledge, pages
Gehlen-Baum, V., Pohl, A., Weinberger, A., and Bry,
F. (2012). Backstage designing a backchannel
for large lectures. In Proceedings of the Euro-
pean Conference on Technology Enhanced Learning,
ucken, Germany (18-21 September 2012).
Goldie, J. G. S. (2016). Connectivism: A knowledge learn-
ing theory for the digital age? Medical Teacher,
Hattie, J. and Timperley, H. (2007). The power of feedback.
Review of Educational Research, 77(1):81–112.
Heller, N. and Bry, F. (2019a). Nudging by predicting: A
case study. In Proceedings of the 11th International
Conference on Computer Supported Education - Vol-
ume 2: CSEDU, pages 236–243. SciTePress.
Heller, N. and Bry, F. (2019b). Organizing peer correction
in tertiary stem education: An approach and its evalu-
ation. International Journal of Engineering Pedagogy
(iJEP), 9(4):16–32.
Heller, N. and Bry, F. (2021). Chapter 14 - human computa-
tion for learning and teaching or collaborative tracking
of learners’ misconceptions. In Caball
e, S., Demetri-
adis, S. N., G
anchez, E., Papadopoulos, P. M.,
and Weinberger, A., editors, Intelligent Systems and
Learning Data Analytics in Online Education, Intel-
ligent Data-Centric Systems, pages 323–343. Aca-
demic Press.
The Future of Higher Education Is Social and Personalized! Experience Report and Perspectives
Henderson, M., Selwyn, N., and Aston, R. (2017). What
works and why? student perceptions of ‘useful’digital
technology in university teaching and learning. Stud-
ies in Higher Education, 42(8):1567–1579.
Hoyt, A., McNulty, J. A., Gruener, G., Chandrasekhar, A.,
Espiritu, B., Ensminger, D., Price Jr, R., and Naheedy,
R. (2010). An audience response system may in-
fluence student performance on anatomy examination
questions. Anatomical Sciences Education, 3(6):295–
Kay, R. H. and LeSage, A. (2009). Examining the benefits
and challenges of using audience response systems:
A review of the literature. Computers & Education,
Keough, S. M. (2012). Clickers in the classroom: A review
and a replication. Journal of Management Education,
Kirkwood, A. and Price, L. (2008). Assessment and student
learning: a fundamental relationship and the role of
information and communication technologies. Open
Learning: The Journal of Open, Distance and e-
Learning, 23(1):5–16.
Kiy, A., List, C., and Lucke, U. (2017). A virtual environ-
ment and infrastructure to ensure future readiness of
data centers. European Journal of Higher Education
IT (EJHEIT), (2017-1).
Li, H., Xiong, Y., Hunter, C. V., Guo, X., and Tywoniw, R.
(2019). Does peer assessment promote student learn-
ing? a meta-analysis. Assessment & Evaluation in
Higher Education, 45(2):193–211.
Mader, S. and Bry, F. (2019a). Audience response sys-
tems reimagined. In Herzog, M. A., Kubincov
a, Z.,
Han, P., and Temperini, M., editors, Advances in Web-
Based Learning ICWL 2019, pages 203–216, Cham.
Springer International Publishing.
Mader, S. and Bry, F. (2019b). Fun and engagement in lec-
ture halls through social gamification. International
Journal of Engineering Pedagogy, 15(2):117–136.
Mader, S. and Bry, F. (2019c). Phased classroom instruc-
tion: A case study on teaching programming lan-
guages. In Proceedings of the 11th International Con-
ference on Computer Supported Education - Volume
1: CSEDU, pages 241–251. SciTePress.
Mader, S. and Bry, F. (2019d). Towards an annotation
system for collaborative peer review. In Interna-
tional Conference in Methodologies and intelligent
Systems for Techhnology Enhanced Learning, pages
1–10. Springer.
Markova, T., Glazkova, I., and Zaborova, E. (2017). Qual-
ity issues of online distance learning. Procedia-Social
and Behavioral Sciences, 237:685–691.
Mazur, E. and Somers, M. D. (1999). Peer instruc-
tion: A user’s manual. American Journal of Physics,
Narciss, S. (2008). Feedback strategies for interactive learn-
ing tasks. Handbook of research on educational com-
munications and technology, 3:125–144.
Nicol, D. (2010). From monologue to dialogue: improv-
ing written feedback processes in mass higher educa-
tion. Assessment & Evaluation in Higher Education,
Oigara, J. and Keengwe, J. (2013). Students’ perceptions
of clickers as an instructional tool to promote active
learning. Education and Information Technologies,
Riazy, S. and Simbeck, K. (2019). Predictive algorithms in
learning analytics and their fairness. In Pinkwart, N.
and Konert, J., editors, DELFI 2019, pages 223–228,
Bonn. Gesellschaft f
ur Informatik e.V.
Slade, S. and Prinsloo, P. (2013). Learning analytics: Ethi-
cal issues and dilemmas. American Behavioral Scien-
tist, 57(10):1510–1529.
Stab, C. and Gurevych, I. (2017). Recognizing insuffi-
ciently supported arguments in argumentative essays.
In Proceedings of the 15th Conference of the Euro-
pean Chapter of the Association for Computational
Linguistics: Volume 1, Long Papers, pages 980–990.
Strickroth, S., Bußler, D., and Lucke, U. (2021).
Container-based dynamic infrastructure for education
on-demand. In Kienle, A., Harrer, A., Haake, J. M.,
and Lingnau, A., editors, DELFI 2021 – Die 19. Fach-
tagung Bildungstechnologien, Lecture Notes in Infor-
matics (LNI), pages 205–216, Bonn. Gesellschaft f
Informatik e.V.
Strickroth, S., Olivier, H., and Pinkwart, N. (2011).
Das GATE-System: Qualit
atssteigerung durch Selb-
sttests f
ur Studenten bei der Onlineabgabe von
Ubungsaufgaben? In Rohland, H., Kienle, A.,
and Friedrich, S., editors, Tagungsband der 9. e-
Learning Fachtagung Informatik (DeLFI), GI Lecture
Notes in Informatics, pages 115–126, Bonn, Germany.
Gesellschaft f
ur Informatik e.V.
Strickroth, S., Striewe, M., M
uller, O., Priss, U., Becker, S.,
Rod, O., Garmann, R., Bott, O. J., and Pinkwart, N.
(2015). ProFormA: An XML-based exchange format
for programming tasks. eleed, 11(1).
van Popta, E., Kral, M., Camp, G., Martens, R. L., and Si-
mons, P. R.-J. (2017). Exploring the value of peer
feedback in online learning for the provider. Educa-
tional Research Review, 20:24–34.
Wenger, E., McDermott, R., and Snyder, W. M. (2002). Cul-
tivating Communities of Practice. Harvard Business
School Press, Boston, USA.
Zheng, L., Zhang, X., and Cui, P. (2019). The role of
technology-facilitated peer assessment and supporting
strategies: a meta-analysis. Assessment & Evaluation
in Higher Education, 45(3):372–386.
CSEDU 2022 - 14th International Conference on Computer Supported Education