Towards a Blended Learning Approach to Teach a Theoretical
Computer Science Module
Leila Silva
1
and Leonor Barroca
2
1
Departamento de Computação, Universidade Federal de Sergipe, Av. Marechal Rondon, s/n, São Cristóvão-SE, Brasil
2
Faculty of Mathematics, Computing and Technology, The Open University, Walton Hall, Milton Keynes, U.K.
Keywords: Blended Learning, Flipped Classroom, Theoretical Computer Science.
Abstract: Theoretical computer science is a difficult subject in the computer science curriculum. Innovations in
teaching and new pedagogic practices have been developing in the last decade but are still far from being
widely applied to computer science. We propose that the teaching of more challenging areas of computer
science can benefit from opportunities created by a blended approach of face-to-face with online teaching
and individual and group activities. We present the design of a Design and Analysis of Algorithms including
innovations in pedagogy, as flipped classroom, problem-based lectures and social learning.
1 INTRODUCTION
In Computer Science teaching, the more theoretical
modules have been known to be hard for students. In
1988, Robins (1988) reported his experience of
teaching theoretical modules in the University of
California: “theoretical computer science has an
awful reputation among undergraduates.... I have
heard many resentful undergraduates describe this
course using terms such as dry, boring, unmotivated,
contrived, impractical, and too abstract.
Interestingly, those very few students (usually those
who excel in the material) describe it as elegant,
challenging, practical, and stimulating.”. He also
reported that in a 50-student undergraduate class
there were 2 or 3 individuals that achieved near-
perfect scores. More than a decade later,
Hamäiläinen (2004) reported that at most a third of
the students that registered for a theoretical module
on computability would pass it. More recently,
Enström (2014) also mentions the challenge of
teaching theoretical computer science and reports
some experience of introducing more interactive
activities to improve the students’ understanding.
From the experience of the first author, in Brazil,
student grades in these modules are lower than in
other modules, and the dropout and failure rates are
usually the highest. In particular, at the Universidade
Federal de Sergipe (UFS), in average, 50% of the
students dropout or fail the theoretical modules. We
do not have statistical analysis to justify this
scenario, but many lecturers who teach these
modules attribute the cause for bad performance to
the poor mathematical background of students.
Despite the importance of theoretical computer
science in the curriculum, there are only a few
studies that apply innovations in pedagogy to this
area, in order to improve motivation, engagement
and performance of students. Recent studies include
Hamäiläinen (2004) using a problem-based approach
to teach computability, Enström and Khan (2010)
introducing lab exercices to teach NP-completeness
proofs and Chakraborty et al. (2011) reviewing the
main initiatives in the use of simulators to teach
automata theory. These works, however, do not
address the entire design of a theoretical module.
The popularization of the use of Information and
Communication Technologies (ICTs) in the last
decades allows for the exploitation of new
pedagogic practices, such as social learning, in
which students change their role away from
information consumers to start engaging in active
cooperation to produce knowledge (Brown and
Adler, 2008; Sharples et al., 2013).
The face-to-face (f2f) model of education has
incorporated some tools commonly used in distance
education, and blended learning has emerged as a
new trend in education. According to Garrison and
Vaughan (2008) blended learning is defined as, “the
thoughtful fusion of f2f and online learning
experiences. [..] f2f oral communication and online
written communication are optimally integrated such
319
Silva L. and Barroca L..
Towards a Blended Learning Approach to Teach a Theoretical Computer Science Module.
DOI: 10.5220/0005483103190324
In Proceedings of the 7th International Conference on Computer Supported Education (CSEDU-2015), pages 319-324
ISBN: 978-989-758-108-3
Copyright
c
2015 SCITEPRESS (Science and Technology Publications, Lda.)
that the strengths of each are blended into a unique
learning experience congruent with the context and
intended educational purpose.” A module designed
to be blended may combine f2f classes, small and
large groups, self-directed learning, communication
between lecturer and students and between students
(Bath and Bourke, 2010). In general, it is possible to
blend time, place, resources and activities. For
example, a module may include f2f and video
lectures, online forums, small group tutorials on-
campus, online and f2f quizzes. Several works have
reported the positive effect of the use of blended
learning and ICT resources (e.g Dziuban et al., 2004;
Rovai and Jordan, 2004; Wang et al., 2008).
Teaching of computer science and, in particular, of
more theoretical areas could benefit from the
opportunities created by the use of blended learning.
We present the design of a theoretical computer
science module using blended learning. We detail
the approach which includes flipped classroom,
social learning and problem-based learning, and
show some illustrative artefacts. We intend to
evaluate our design to answer questions such as:
How can blended learning help with
motivation and learning of theoretical
computer science subjects?
Which methods, practices, tools and
resources can be used in a blended module
in theoretical computer science?
Which is the role of social learning in the
learning of theoretical computer science?
Which kinds of open educational resources
(OERs) can be used to improve motivation
and understanding of hard topics?
What is the impact of using flipped
classroom vs. strictly f2f classes?
The paper is organised as follows: section 2
introduces the design approach used; section 3
introduces the proposed module in the context of the
computing curriculum at the UFS; section 4 reviews
related work, and section 5 presents the discussion
and directions for further research.
2 METHODS
We combine the approaches of Bath and Bourke
(2010) and Conole (2010). The former gives us the
process, within which the conceptual views
proposed by the later are elaborated in an interactive
and incremental way. The process comprises five
phases: planning, design, development,
implementation, review and improvement.
In the planning phase we define: aims and
learning objectives of the module, structure and
timetable, materials and resources, teaching and
learning activities (including communication and
collaboration between students and lecturer and
between students), teaching strategies (for example,
which part of the course should be online or f2f),
assessment, and student feedback.
In the design and development phases we detail
how learning objectives, teaching and learning
activities, and assessment are integrated to enable
the lecturer to judge constructive alignment (Biggs
and Tang, 2011). This means that we evaluate
whether resources and learning and teaching
activities support students achievement of the stated
learning objectives. We also judge if assessment is
consistent with the activities and objectives. We
decide on the resources to be used and on the
workload. We detail, for each learning objective,
which activities take place f2f and online and decide
on the balance of the types of activities undertaken
by students. We use Conole’s (2010) pedagogy
profile that defines the following types: assimilative
(attending and understanding contents), information
handling (gathering and classifying resources or
manipulating data), adaptive (using of modelling or
simulation software), communicative (dialogic
activities, such as group based discussions),
productive (construct an artefact) and experiential
(practising skills in a particular context or
undertaking an investigation).
For the implementation phase, when the module
is taught, we prepare a welcome orientation for
students explaining the blended approach used.
In the review and improvement phase we collect
and analyse feedback on different aspects of the
module (e.g. content, activities, assessments, etc).
This feedback provides an opportunity to review
different aspects of the module and to reflect on
improvements for future instantiations.
3 PROPOSED MODULE
3.1 Background
Design and Analysis of Algorithm is a second year
module of the BSc in Computer Science at UFS.
Until now the module has been offered strictly f2f,
with 50 hours of lectures and 10 hours of
assessment. It is expected that students engage in at
least six hours of extra-class reading and exercises
per week. Assessment comprises four tests and two
exams. The lecturer is assisted by at most two more
CSEDU2015-7thInternationalConferenceonComputerSupportedEducation
320
senior students, here called assistants, who have
passed this module with good grades. The size of the
classes ranges from 30 to 55 students each semester.
The success rate is around 50% or less. In general, in
a 0 to 10 scale, only up to 10% of the students who
pass the module have grades above 8; the majority
of grades range between 5.5 and 6.5. This scenario is
not common in other advanced modules, and even in
comparison with first years modules the success rate
is low.
Feedback questionnaires applied in some classes
reveal that most student complains are about the
difficulty of the subject, the need for more problem
solving classes, and for better integration between
the theoretical content and real world problems.
Students recognise that this module, unlike others,
requires continuous dedication to study and that it is
not possible to pass a theoretical module only
studying for the tests.
3.2 Module Design
In what follows we give an overview of the module
design, according to the approach described in
section 2.
3.2.1 The Planning Phase
An overview of the module is expressed by the
Module Map View (Conole, 2010). This artefact
enables lecturers to think about, and share, the
design of the module considering the following meta
aspects: Guidance & Support, Contents & Activities,
Reflexion & Demonstration and Communication &
Collaboration. Guidance & Support include some
elements such as module structure, timetable and
human resources. Contents & Activities include the
topics and activities of the module and the materials
used. Reflexion & Demonstration define how
internalization and reflection is carried out.
Communication & Collaboration list the techniques
and resources that support the interaction between
students and lecturer or between students. In
addition, a module summary and key words
indicating the pedagogical approach are provided at
the beginning of artefact. The Module Map View
has a worksheet format, but due to space restrictions,
here we present its content in a textual form.
Module Summary: 2
nd
year course; 4 credits over 15
weeks; 5 blocks covered by 12 theoretical f2f
lectures, 11 theoretical on-line lectures, 10 problem
solving classes, 1 project guideline class, 4
assessment classes, 2 project presentation classes, 1
consolidation and feedback class.
Key Words: Blended Learning; Flipped Classroom.
Guidance & Support: module guide; study calendar;
study planner; VLE, Google, Google docs, social
learning tools, programming environments and
tools; one lecturer and two assistants.
Contents & Activities: Blocks - (1) Introduction to
Complexity Theory, (2) Searching and Sorting, (3)
String Processing, (4) Geometric Algorithms, (5)
Problems solved by Dynamic Programming and
Backtracking techniques; Activities - assimilative
(f2f and online lectures); communicative (problems
to discuss and solve in pairs or in group); productive
(problems to implement in group, presentation of
solutions); experiential (analyse and solve a real
problem applying studied algorithms); Resources -
books; lecture notes; lecture videos available in
VEDUCA (www.veduca.com.br); educational
resources from the Web; problems in programming
environments; learner-generated presentations.
Reflexion & Demonstration: Assessment -
assimilative (quizzes and tests), communicative and
productive (problems and project), experiential
(project); self-evaluation questions; feedback
questionnaire.
Communication & Collaboration: problem solving
in pairs; group discussion and implementation of
problems; group project; individual or tutor groups
led by lecturer and assistants; email and news trough
the VLE; chats between students and assistants
through social learning tools; online materials and
notes; f2f classes.
3.2.2 The Design and Development Phases
In this phase we use artefacts that relate the learning
objectives to the activities in the module. Here we
show some of these artefacts: the module flow gives
an overview of the sequence of the activities within
the module; the blended learning design worksheet,
adapted from (Bath and Bourke, 2010), shows the
learning objectives, how they are assessed, and the
teaching and learning activities and resources to
achieve those objectives; and the pedagogy profile
(Conole, 2010) shows the distribution and weight of
the different types of activities.
Module Flow
The module includes f2f and online lectures. F2f
lectures have two hours of duration and can either
consist of the explanation of a topic or be a problem
solving class. Online lectures explain a topic through
slides, commented with audio recordings. Each
problem solving class relies on the contents of
TowardsaBlendedLearningApproachtoTeachaTheoreticalComputerScienceModule
321
online lectures, implementing the flipped classroom
model, in such a way that the theory learned extra-
class is applied and discussed in f2f-classes.
The slides for each lecture (f2f and online) are
delivered to students in the VLE. These slides are
elaborated by the lecturer and may integrate OERs.
They may include also a list of recommended
exercises, chapter readings and selected articles.
Some videos available in VEDUCA may be
recommended. VEDUCA is a repository of video
lectures from well recognised universities translated
to Portuguese.
A variety of assessments is considered. To
stimulate continuous learning, quizzes are applied. A
quiz consists of short questions to test understanding
of the main concepts taught in the previous lecture.
To improve the link between theory and practice, a
supervised group project has been introduced. It
consists of finding and solving a real problem using
some of the algorithms learned in the module. Four
tests are applied; they test problem solving skills
using the techniques learned in the module.
The module is divided in five blocks. The first
provides the basis for the other blocks and requires
more mathematical background. Experience shows
that students need more interaction and support to
understand the topics of this block and to be
motivated for engaging in the module’s activities.
Therefore, the block is taught in f2f classes. Some
classes include quizzes, and a test is applied at the
end of the block. Guidelines about the project are
given to students at the end of the first block.
Blocks 2 to 5 are taught with lectures and
problem solving classes. As problem solving classes
require the knowledge of online lectures, students
are required to study the topics covered in online
lectures before attending problem solving classes.
These are designed in pairs, in such a way that the
solutions achieved in one class of a pair will be
discussed in the next class. In a problem solving
class the lecturer starts by clarifying any doubts
from the corresponding online lecture. After that, the
students answer a quiz, individually. Then, the
students are divided in groups of 4 and a different
problem is assigned to each group. Each group is
subdivided in pairs to work on a solution. Finally,
the whole group discusses the best solution in the
group. In this process, students are assisted by the
lecturer and assistants. The group is asked to
implement the solution outside the class and to
prepare a short presentation, using slides, to discuss
it in the next class, with all students and lecturer.
Students are also advised to solve the other groups’
problems before the next class, to take an active part
in the discussion of the different solutions. All
solutions will be available through the VLE after the
second class, for further discussion using social
learning tools. A test is applied at the end of blocks
2, 3 and 5. Students are required to develop their
project before the end of block 5. They have extra-
class support given by the lecturer and assistants for
the project. The project is presented after block 5.
Blended Learning Design Worksheet
Here we illustrate the content of the worksheet by
taking as example block 2, which is taught in a
flipped classroom approach and is a good
representative of the module design.
Example
Learning Objective: understanding and applying the
main sorting algorithms.
Ways of Assessing the Objective: f2f quizzes, pair
and group exercises in f2f classes and extra-class
activities, test. The purpose of the quizzes is to
check the understanding of the main concepts of the
sorting algorithms, whereas the ability of applying
these algorithms in diverse situations is verified in
exercises and block test.
Teaching Activities: a f2f class about a robust
sorting algorithm; online lectures about other sorting
algorithms; opportunities to discuss the main
difficulties from the online lectures in f2f classes;
problem solving f2f class about sorting algorithms;
recommended readings; a list of exercises for home
work; feedback on assessment and students’ work;
suggestions for online lectures on the subject from
platforms such as VEDUCA; preparation of the
assistants for the evaluation of students’ work.
Assistants play also a teaching role helping with:
tutoring students in chat rooms and f2f; preparing a
practical f2f class on the use of programming
environments; selecting problems to explore in
problem solving classes; helping the lecturer in
problem solving classes, and with marking.
Learning Activities: study the main sorting
algorithms, propose solutions for recommended
exercises, complete assessments, work in pairs and
in a group of 4 to solve and implement problems,
prepare a short presentation of solutions achieved,
discuss solutions of problems, watch videos about
the subject from other universities, access assistants,
lecturer and other students to clarify doubts.
Helpful Resources: books and OERs, programming
environments, VLE and social learning tools.
The Pedagogy Profile
The pedagogy profile (Figure 1) gives an overview
of the distribution of learning activities: assimilative,
CSEDU2015-7thInternationalConferenceonComputerSupportedEducation
322
information handling, adaptive, communicative,
productive and experiential. The balance between
productive and assimilative activities, shows the
integration between practical and theoretical
activities, one goal of the module conception.
Figure 1: The Pedagogy Profile.
3.2.3 The Implementation and Review and
Improvement Phases
The blended module proposed is under development
and will be implemented in the first semester of
2016. The authors are currently preparing the
module materials.
An experiment will be conducted to validate the
approach. The same content and similar assessment
will be applied in two classes, one following a
blended approach and the other a traditional
approach. We want to identify whether students
benefit from the innovations introduced. The
dropout rate and students’ grades, in each block and
in the whole module, will be compared.
A feedback questionnaire will be applied at the
end of each block and of the module, to measure
students’ satisfaction concerning various aspects of
the module. For example, we intend to analyse the
students perception of workload for each activity in
the pedagogy profile and the consistency between
assessment and learning objectives. We intend to use
feedback from the questionnaires to support future
iterations of the module.
4 RELATED WORK
Blended learning needs to be strategically planned to
be adopted in the whole curriculum or in a specific
module (Mortera-Gutiérrez, 2006; Oblinger, 2006).
However, there is no well-established procedure to
design a blended learning program (Oliver and
Trigwell, 2005) and different researchers have
suggested different approaches based on five
blending dimensions: online and offline learning,
self-paced and collaborative learning, structured and
unstructured learning, custom and off-the-shelf
content and prior and on demand support (Singh,
2003; Garisson and Vaughan, 2008; Dziuban et al.,
2004; Larson and Murray, 2008).
Blended learning is being used successfully in
several areas of higher education (see for example,
Dziuban et al., 2004; Rovai and Jordan, 2004;,
Holley and Dobson, 2008), as well as in training
programs (Moe and Rye, 2011). In Computer
Science, blended modules have been recently
experimented with. For example, Alonso et al.
(2011) report an experience of a blended module of
Program Development Models, in which the
approval rate is significantly higher than in the
strictly f2f module. Similar results have been
achieved by Deperliogli and Kose (2013), in the
context of Data Structures. Marin and Pascual Nieto
(2012) introduced a free-text score system to
encourage students to study core concepts of an
Operating System module after classes. Gannod et
al. (2008) adopted the flipped classroom approach in
the design of a Software Engineering module. Day
and Foley (2006) combined successfully video
lectures with f2f exercise classes in a module of
Human Computer Interaction. Nevertheless, we are
not aware of any blended learning approach to teach
Design and Analysis of Algorithms as proposed
here.
5 CONCLUSIONS
In this paper we propose the introduction of a
blended approach to the teaching of the module
Design and Analysis of Algorithms. In the design of
the module, we introduce some innovations in
pedagogy such as flipped classroom, social learning
and problem-based learning. OERs, such as video
lectures, are used to widen the learning
opportunities.
New materials are being identified/adapted/
prepared to help with more difficult topics. The
engagement of students in activities is stimulated by
the introduction of quizzes and problem solving
classes. The introduction of a supervised project is
intended to diminish the gap between theory and
practice. The problem solving classes and the
supervised project are intended to improve students’
collaboration and communication skills.
The proposed module is still under construction
and will be implemented in the first semester of
2016 at the UFS. We expect that the change in the
TowardsaBlendedLearningApproachtoTeachaTheoreticalComputerScienceModule
323
module’s dynamics, mixing online and f2f lectures,
introducing new forms of assessment and
opportunities for social learning will improve the
dropout rate and grades.
The immediate future work is to conduct an
experiment in order to compare the traditional and
the blended approaches in the learning of the same
module. As theoretical computer science is a
challenging subject, this case study should be
regarded as a proof of concept for the applicability
of blended learning. We also hope to contribute to
the introduction of pedagogic innovations in the
Computer Science curriculum.
ACKNOWLEDGEMENTS
This work was partially supported by the National
Institute of Science and Technology for Software
Engineering (INES), funded by CNPq, grant
573964/2008-4.
REFERENCES
Alonso, F., Manrique, D., Martinez, L., Vines, J M., 2011.
How Blended Learning Reduces Underachievement in
Higher Education: An Experience in Teaching
Computer Science. In IEEE Transactions on
Education, 54(3): 471-478.
Bath,D., Bourke, J., 2010. Getting Started with Blended
Learning. Griffith Institute for Higher Education.
Technical Report.
Biggs, J., Tang, C., 2011. Teaching for quality learning at
university (Society for Research in Higher Education),
The Open University Press, Milton Keynes, 4
th
Edition.
Brown, J. S., Adler, R. P., 2008. Minds on Fire: Open
Education, the Long Tail, and Learning 2.0. In
Educause Review, 43(1):16-32. Educause.
Chakraborty, P., Saxena, P.C., Katti, C.P., 2011, Fifty
Years of Automata Simulation. In ACM Inroads,
2(4):59-70.
Conole, G., 2010. Learning design- making practice
explicit. In ConnectEd Design Conference, 1-31.
Day, J. A., Foley, J. D., 2006. Evaluating a Web Lecture
Intervention in a Human Computer Interaction Course.
In IEEE Transactions on Education, 49(4): 420-431.
Deperliogli, O., Kose, U., 2013. The Efectiveness and
Experiences of Blended Learning Approaches to
Computer Programming Education, In Computer
Applications in Eng. Education, 21(2): 328-342.
Dziuban, C. D., Hartman, J. L., Moskal, P. D., 2004.
Blended Learning. EDUCAUSE Center for Applied
Research Bulletin.
Enström, E., Kann, V., 2010. Computer Lab Work on
Theory. In Proc. ITiCSE, 26-30.
Enström, E., 2014. On Difficult Topics in Theoretical
Computer Science Education, Doctoral Thesis, KTH
School of Computer Science and Communication,
Stockholm.
Gannod, G. C., Burge J. E., Helmick, M. T., 2008. Using
Inverted Classroom to Teach Software Engineering. In
Proc. of ICSE Conf., 777-786.
Garrison, D. R., Vaughan, N. D., 2008. Blended Learning
in Higher Education: Framework, Principles, and
Guidelines. John Wiley & Sons, Hoboken, 1
st
edition.
Hämäiläinen, W., 2004. Problem-based Learning of
Theoretical Computer Science. In 34
th
ASEE/IEEE
Frontiers in Education Conference, 20-23.
Holley, D., Dobson, C., 2008. Encouraging students
engagement in a blended learning environment: The
use of contemporary learning spaces. In Learning
Media and Technology, 33(2): 139-150.
Larson, R., Murray, M., 2008. Open educational resources
for blended learning in high schools: overcoming
impediments in developing countries, J Asynchronous
Learn Netw, 12(1): 1-19.
Marin, D. P., Pascual Nieto, I., 2012. A Case Study on the
Use of Blended Learning to Encourage Computer
Science Students to Study. In Journal Sci. Educ.
Technol., 21: 74-82.
Moe, C. E., Rye, S. A., 2011. Blended Learning:
communications, location and work-life practices.
Educational Media International, 48(3):165-178.
Mortera-Gutiérrez, F., 2006. Faculty best practices using
blended learning in e-learning and face-to-face
instruction. International Journal on Elearning,
ProQuest Education Journals, 5(3): 313-337.
Oblinger, D. G., 2006. Space as a change agent. In D. G.
Oblinger (Ed.), Learning spaces (pp. 1.3). Educause e-
book.
Oliver, M., Trigwell, K., 2005. Can ‘blended learning’ be
redeemed? E-learning, 2 (1): 17-26.
Robins, G., 1988. Teaching Theoretical Computer Science
at the Undergraduate Level: Experiences,
Observations and Proposals to Improve the Status
Quo, Technical Report, UCLA CSD, vol. 880063.
Rovai, A. P., Jordan, H. M., 2004. Blended Learning and
Sense of Community: a Comparative Analysis with
Traditional and Fully Online Graduate Courses. In Int
Rev of Research in Open and Distance Learning, 5(2).
Sharples, M., McAndrew, P., Weller, M., Ferguson, R.,
Fitzgerald, E., Hirst, T., Gaved, M., Smith, J., 2013. In
Innovating Pedagogy. The Open University
Innovation Report 2.
Singh, H., 2003. Building effective blended learning
programs. Educ Technol Mag, 43(6):51-54.
Wang, F. L., Fong, J., Choy, M., Wong, T., 2008. Blended
Teaching and Learning of Computer Programming. In
Int. Conf. on Web-based Learning, LNCS 4823, 606-
617, Springer.
CSEDU2015-7thInternationalConferenceonComputerSupportedEducation
324