Mixed competitive -collaborative approach in [621925]

Mixed competitive -collaborative approach in
education (January 2015)
Tănase Gula, Cristina Grosu, Diana Nănuți
Computer Science and Information Technology
University “Politehnica” of Bucharest, Romania
[anonimizat], [anonimizat], [anonimizat]

Abstract — Nowadays , all university level programs are obligated
to prepare students for professional employment while
simultaneously providing the academic rigor c onsistent with
university level study. To provide computer scientists with good
materials and interesting topics in a class does not necessarily
mean that their education is of a high quality, students need to be
motivated and evolve skills needed in a rea l-life employment.
Social skills, teamwork, collaboration and competition are
valuable aspects they should know in other to become
professionals.
Therefore, it is critical that professional studies students develop
the requisite skills to succeed in a team -oriented environment
within the competitive background of commercial industry. This
paper presents a study with intention to improve education of
computer science students in employment -like environments. The
study utilizes experience with mixed competiti ve and collaborativ e
learning in education, applied to a programming course.
Index Terms —competitive learning, collaborative learning,
cooperative learning, computer science, education.

I. INTRODUCTION
OMPUTER science has multiple disciplines that are
continuously changing and evolving. Since this domain is
so competitive so has to be the education the students
receive. Competition is evident throughout our society, lives
and work. In order to succeed, students should have social
skills, communication cap abilities, competitive nature and
ability to collaborate in a team besides knowledge of
technologies, techniques, models and processes. All these
aspects are important when a student: [anonimizat],
so the goal of a teacher is to prepare profes sionals that can
succeed in the employment.
The goal of this paper is to improve education of computer
science that will help them with the adaptation towards their
future career. It is a known fact that the traditional education
provides a good overview of techniques and technology, but does not necessarily focus on the related aspects that are
essential for becoming a professional [1]. The purpose of this
paper is to present a mixed collaborative -competitive method
that can help students acquire suppleme nt knowledge in their
field of interest.
There are some differences between the two approaches in
education. Cooperative learning may be contrasted with
competitive learning, in which students work against each other
to achieve a good grade and only some of them succeed and
individualistic learning, in which students work independently
to achieve learning goals unrelated to those of other students.
Within competitive situations, individuals seek outcomes that
are beneficial to themselves and detrimental to others. The
student: [anonimizat] "they can obtain their goals if and only if
the other students in the class fail to obtain their goals." [2].
Unlike the competitive approach, the collabo rative approach
refers to students that have a common objective, a common task
or where each individual depends on and is accountable to each
other.
Both of these techniques can be found in the workplace of
our times. On one hand, collaborative activities for school
projects for example can prepare students for the experience of
working in a team in their professional life. On the other hand
competitive learning has the potential to engage students in the
learning process. The competition between students c an inspire
even the ones not interested by the subject at the beginning and
also encourage those who are not prone to be competitive.
Furthermore, various benefits have been observed by using
one of these learning methods, being either collaborative or
competitive; for example:
 Improve depth of thought by comparing different ideas,
as opposed to passive listening.
 Promote a better student -student and student -faculty
interaction.
 Encourage improvement and innovation in student C

involvement or teaching environment and techniques.
 Promote a better attendance and a higher level of
achievement.
Cooperative learning has several advantages, such as:
 Offers the opportunity of learning how to work together
with others, as part of a team.
 Allows the students to define their favorite learning
style, by working together with peers, rather than
working individually.
 Acknowledges that goals in life can be achieved in
cooperation with each other.
 A new enjoyable overview upon school and learning for
learners.
 Students learn to assist each other in the school and team
setting, which can be extremely helpful later on in their
professional career.
This paper is organized as follows: section 2 describes the
background of computer science, competitive and collaborative
learning and previous research in the approach, while the
methodology and supporting framework are presented in
Section 3.
In Section 4, we discuss our findings and present student
opinions for this methodology, based on empirical and
questionnaire results.
Finally, in Section 5, overall conclusions are taken and future
work is presented.

II. PREVIOUS WORK
Computer science spreads over multiple disciplines such as
theoretical computers science, algorithms, design patterns,
databases, software engineering and many others. Education in
these areas often involves math, graph theories, mathematical
logic and statistics.
Since the field has so many different areas, it is important to
have a flexible education and evaluation system, in order to
emphasize the particularities of each subject.
Having this in mind, we have to think about both traditional
and alternative pedagogical methods that are appropriate for
different learning styles.
For computer science courses a lecturer traditionally follows
a table -driven student evaluation [14]. The teacher uses
summary tables to collect the grades for each student . For
example , considering the following points ’ distribution for the
project’s grading, shown in Table 1, and the project evaluation

with its tasks distri buted as well, like in Table 2, then the
summary table will look like Table 3.
The summary table centralizes the data and can be used for
statistics [15] (e.g. to compute grades' distribution, average
etc.).

This often statistically distributes students in groups for
grading according to the Ga ussian distribution. This method of
evaluation has always had the experts wondering if this is the
most suitable approach or not. Hence, there are several learning
styles proposals and mixed applications described in literature.
Johnson and Johnson [2] ref er a close relationship between
learning styles and attitudes towards learning, including
motivation to learn, involvement in learning activities, attitudes
towards each -others and self -efficacy. For students, it is easier
to learn when the information is presented in a way that matches
their learning style. Moreover, independently of the area of
knowledge, recent innovations in pedagogical techniques have
led to the in troduction of new instructional methods in the
classroom. These innovations include, for example, educational
and on -line games [5, 6], asynchronous instruction (e.g., email, No. Task Points per task
1 Individual strategy 40
2 Team strategy 30
3 Cooperation 10
4 Solution complexity 20

Table 1 – Project Grading System
No. Task Points per task
1 Exam 25
2 Lab 10
3 Course presentation 15
4 Project 50

Table 2 – Overall Grading System
Student Final
Grade Exam
(25%) Lab
(10%) Course
(15%) Project
(50%)

Table 3 – Summary Table

electronic bulletin boards and podcasting) and computer -based
teaching [7].
Having all these instruments at their disposal, teachers should
select the most appropriate le arning approach for their subject,
in accordance to their previous experiences, their personal
preferences, and concepts they want to pass through.
A good approach are the competitions organized within
schools or among them, when students are organized in teams
and are encouraged to collaborate between their group members
and to compete with the other ones, in order to achieve a well –
known goal. In our university there are several contests that
stimulate this approach.
For example the ACM International Col legiate Programming
Contest, the programming competition organized among the
universities of the world. The students can participate alone or
as part of a team with maximum 3 members and have to solve
algorithmic problems within a given time. They are allo wed to
use one laptop for coding and one laptop to read the
requirements of the problem. Also, to get a higher score you
need not only to solve the problem correctly, but to be the first
team who solves it. This way, the students have to collaborate
within their team to correctly understand the requirements and
establish the best algorithm for the given problem and the most
efficient way of implementing it.
The competitive attitude is encouraged within disciplines as
well. For example, a homework consisted in implementing an
AI that plays quantic Tic -Tac-Toe in Prolog, and the first 10
students who defeated the bots received 20% bonus points.
Also, every year our university organizes for freshmen a
scavenger hunt, where they form teams of 5 members and are
given a map of the campus and hints in order to find all the
items. With no doubt, there is a competition among students and
also collaboration within their teams, which involves student –
to-student interaction.
The work focuses on five conditions that sup port cooperative
learning as more productive than competitiveness. These
conditions are:
● Perceived positive interdependence
● Face -to-face interaction
● Personal responsibility to achieve the group's goals
● Use of the relevant interpersonal and small -group
skills
● Group processing of current functioning to improve
the group's future effectiveness in mathematics.
Results showed that in this kind of activity, the cooperative
learning strategy was more effective than the competitive
strategy. An ex ample of this kind of results is an experiment
made for an organic chemistry course, on a period of three
years, using peer -led team learning, maintaining the teacher, the grading system and the examination part constant. The results
of this study were tha t: “On average, the workshop students
significantly outscored their traditionally -taught counterparts
on individual course exams, final course grades, retention in the
course, and percentage earning the minimum acceptable grade
of C– for moving on to the s econd semester organic chemistry
course.” [ 16]. Please see the reference for more examples of
different cooperative strategies results.
There is also a result that supports keeping the team in a small
size, mainly because it impacts individual accountabili ty [1]. In
this case, it is highly important to let students share their
findings, which also helps them develop their social skills.
Study also shows that strategies are given to collaborative
environment where the focus is on team role assignment [1].

III. PROPOSED APPROACH
Above all, competition can be seen as “a social process that
occurs when rewards are given to people based on how their
performances compare to the performances of others doing the
same task or participating in the same event” [8]. As any other
social process, competitive learning holds both positive and
negative aspects. Researchers found that competition had a
positive impact on performance goals and learning motivation
in the classroom [13].
Also, it has been stated that “learning by l osing” was a
valuable process for students preparing themselves for
professions where working under pressure was necessary [9].
On the other hand, competitive learning brings in our attention
some handicaps identified in literature, like high anxiety level s,
self-doubt, selfishness and the interference with the capacity to
solve problems [2]. Nonetheless, these issues can be attenuated
if one establishes different competing environments,
introducing collaborative learning and supporting collaboration
for a common set of goals and objectives. In contrast to
competitive learning, collaborative learning, has been defined
as “a social process through which performance is evaluated
and rewarded in terms of the collective achievement of a group
of people working t ogether to reach a particular goal” [8]. As
well as competitive learning, collaboration has both positive
and negative aspects.
“Deutsch suggested that collaboration embodies positive
interdependence, supporting students to develop themselves in
a collabo rative environment. On the negative aspects of
collaborative learning, we identify there is the possibility of an
uneven engagement in learning by individual students within a
group, translated into uneven contributions for the group's final
grad to relate d work.” [10]

IV. CASE STUDY
The next pages present two project management approaches:
anarchical and competitive -cooperative, in order to emphasize
the pros and cons of each one, but maintaining the focus on
competitive -cooperative approach.
On one hand, thi s year, the approach tried for the Software
Project Management discipline was the anarchical approach
and consisted in organizing the students in teams of 15 -19
members, depending on the laboratory class they attended to
[11]. The task was to provide a sol ution for a game project with
given rules.
The game environment is a board of known L x L
dimensions, as you may see below in Fig.1, where L represented
the maximum number of members a team could have plus 1.

The players are pawns, graphically represented by colored
circles and by their names. Also, each team has two attributes:
color and luminosity (e.g. as you may see in the legend of Fig.1,
Team A has the color red and its luminosity is black). The
players’ pawns keep the l uminosity from their original teams
and change the color of the team for each round. For the pawns
movement, each student has to write his own code, in order to
implement two strategies: Team strategy – established within
the team, by collaborating with th e other team members and
applied when playing in their own team.
 Individual strategy – different for each player and
applied when they are on another team.
A game round involves 4 teams competing one against
another, distributed along the edges of the boa rd, and chosen
randomly and taking into consideration that at the end of the
game all have played the same number of rounds. Furthermore,
for each round, the members’ shuffle is made so that at least
half of the original team members remain in the initial team.
The game rules were simple. A pawn can move in four
directions, it can move up, down, left, rig ht, but not diagonally and only one cell at a time.
The teacher also had to provide a server used for team
communi cation and for movement purpose.

The team had different tasks, but the most important ones
were:
 to establish the team strategy;
 to select someone to represent the team in the
negotiation with the other teams to establish the
communication messages.
In order to win the game, at least one pawn has to reach the
opposite edge of the board.
On the other hand, last year, the project consisted in voting
based image binarization, using a competitive -collaborative
approach. The project purpose was the development of an
“Image Binarization System” (IBS ), which consisted in two
components:
 “Binarization Algorithm Module” (BAM) – an
executable which produces an output binary image
from a continuous -tone image.
Fig. 1 – Game board

Fig. 2 – Documen t images
(a) Representative image from DIBCO -2011 dataset;
corresponding binarization result produced by the method
(b)Messaoud et al. (2011); (c) Su et al. (2010); (d)
Ntirogiannis et al. (2009); (e) Howe (2011); (f) Gatos et al.
(2008).

 “Voting Binarization Algorithm Module” (VBAM) –
makes a binary image [12] from independent BAM
results.
This can be used in the process of image document analysis,
which involves different operations that are done on simplified
black and white versions of the original images – please see the
examples from Fig. 2 .
The students from each laboratory cla ss were divided in
teams of around 3 members, so that were 4 teams per class. The
work was divided as follows:
 3 teams worked on BAMs – each team developed a
BAM;
 One team developed the VBAM, based on the 3 BAMs
mentioned above and it is responsible of the
synchronization between teams.
To sum up, every team had three basic activities: research,
development, test and their team members were strongly
encouraged to switch their roles.
Furthermore, in both years was organized a contest between
the teams. Last year, the subgroups competed one against
another with the VBAM results, all BAMs competed as well,
and the top N were selected, and all VBAMs, to select the best.
For a bonus, the students could work at the end of the semester
to make the selected VBAM wor k with the top N BAMs. This
year, using the server, the teams played one against the other.
The grading part for the entire course is represented in
Chart1:

It is easy to observe that the project grade was the same
amount of points from the final score in both years, even though
small differences appeared for the others.
V. RESULTS
Using the anarchical approach, the project results were not
that good. The first impediment was the server, which was not
fully implemented, nor f unctional, therefore the evaluation
could not be done as wanted. Also, the students’
implementation of the clients were simplified or non –
functional, since the actual contest was cancelled due to this
technical issue. Opposite to that, another benefit of t he
competitive -collaborative approach is that the failure of a team
in the final functionality does not produce the failure of the
entire project.
The game strategies for this year project were various, from
paths of movement to alpha -beta pruning algorith ms. A
difficulty was found also in establishing a team strategy, since
the teams were large, of 15 -19 members which had to establish
a common point of view, compared to the competitive –
collaborative approach were the small teams were of 3
members, and the larger ones of 12.
Also, in the competitive -collaborative approach, it is easier
to synchronize the teams, since one team was responsible with
the management, whilst in the anarchical approach there was no
one with this role.
The grading part is “fair and square” for the
competitive -collaborative approach, with a balance between
teamwork and individual activity. The anarchical way focuses
more on the individual contribution and efficiency in team’s
strategy.
For the last year project, using the competitive -collaborative
approach, students provided efficient implementations of the
algorithms. Therefore, the grade received for their projects were
high – please see Chart 2, containing the distribution of grades
for all 57 course participants.

Chart 2 – Project Grades Distribution (2013 -2014)
0246810121416
7.6 7.92 8.73 8.8 9 9.1 9.3605 9.4 9.7 10Project Grades Distribution (2013 -2014)
Chart 1 – Points per task

0102030405060
Exam Presentation Project LabPoints per task
2013-2014 2014-2015

For this year project, there was a higher number of
participants than last year, with a total of 84 participants , with
27 more than last year – see Chart 3, and a better view of the
results is shown in Plot 1.

The minimum grade with the competitive -cooperative
approach was 7.6 while with the anarchical approach is 4. 14.
Also, last year the higher grade was 10, while this year was 9.4.

Therefore, the competitive collaborative approach provides a
better environment and student interaction for project
management
VI. FUTURE WORK
The future work will consist in elaborating an online
questionnaire which will be given to the students t o be
completed. Afterwards, the students’ feedback responses will
be analyzed, so that it can be conclude d which approach they
liked more and found more useful. Based on this results, the
next projects can be implemented more efficiently.
An interesting idea would be to try several approaches in the
same year of study, with several teams (e.g. each laboratory
class team to use a different approach) and compare the results
obtained by each of them for the same project and grading
system.
Also, with smaller tasks and more milestones, it would be
easier to manage the project’s risks, by providing a review for
each student (e.g. what needs to be improved, what is good/bad,
what needs to be reworked etc.). This way, the students will pay
more attention to the i mportant details of the project and it will
be easier for them to make a progress in the p roject
development. Another benefit is that the grading part for the
final project would be easier, but this requires, of course, more
time, work and involvement.
Furthermore, in the future, in order to minimize the risks and
impediments, a good idea would be to work on the prerequisites
of the proj ect and have them ready in time, since for this year
project the grades were affected by this part as well.
To sum up, th ere are plenty of perspectives to improve
projects management, but this depends not only on the chosen
strategy, but on the human individuals as well.

VII. REFERENCES
[1] Tomas Cerny, Bozena Mannova “Competitive and Collaborative
Approach toward a More Effective E ducation in Computer Science”,
Contemporary Educational Technology, 2011, pp. 163 -173
[2] D. W. Johnson and R. T. Johnson. “Learning together and alone:
Cooperative, competitive and individualistic learning.” 5th edition.
Needham Heights, MA: Allyn & Bacon, 19 98.
[3] D. A. Kolb and A. Y. Kolb. “Learning styles and learning spaces:
Enhancing experiential learning in higher education.” Academy of
Management Learning & Education, 4(2), pp. 193 -212, 2005.
[4] R. F. McNergney, J. Herbert and R. E. Ford. “Cooperation and
competition in case -based teacher education.” Journal of Teacher
Education, 45(5), pp. 339 -345, 1994.
[5] O. Shabalina, P. Vorobkalov, A. Kataev and A. Tarasenko. “Educational
games for learning programming languages.” Third International
Conference "Modern (e -) Learning" (MeL 2008), Varna, Bulgaria, June –
July 2008.
[6] R. Neves Madeira, J. Sousa, V. Fernão Pires, L. Esteves and O. P. Dias.
“A mobile and web -based student learning system.” WCES -2009 Social
and Behavioural Sciences Procedia Journal, 2009.
[7] G. Catarino , J. Campos, R. Neves Madeira and C. Luz. “b -3m Learning –
A blended multimedia system for mathematics learning. Proceeding s of
International Conference m -ICTE2006, Vol II, pp. 1181 -1185, Formatex,
November 2006.
[8] J. Coakley. “Sport in society: issues and c ontroversies.” 4th edition. S t.
Louis: Times Mirror/Mosbey College, 1994.
Chart 3 – Project Grades Distribution (2014 -2015)

00.511.522.533.544.5
4.14
4.6
5.22
5.41
5.55
5.64
5.88
6.01
6.35
6.44
6.67
6.9
7.02
7.17
7.4
7.52
7.64
7.76
7.87
7.99
8.11
8.27
8.4
8.48
8.63
8.9
9.12
9.22
9.33Project Grades Distribution (2014 -2015)

Plot 1 – Grades Comparison
0246810121416
4 5 6 7 8 9 10Count
Grades2013-2014
2014-2015

[9] J. Dettmer. “Competition photography… learning by losing.” PS A
Journal, 36, June 2005.
[10] M. Deutsch. “A theory of cooperation and competition.” Human
Relations, 2, pp. 129 -152, 1949.
[11] http://elf.cs.pub.ro/mps/wiki/proiect/descriere
[12] http://en.wikipedia.org/wiki/Binary_image [13] S. Lam and P. Yim and J. Law and R. C heung. “The effects of competition
on achievement motivation in Chinese classrooms.” British Journal of
Educational Psycology, 74(2), pp. 281 -296, 2004.
[14] http://en.wikipedia.org/wiki/Evaluation
[15] http://elf.cs.pub.ro/mps/wiki/catalog
[16] http://www4.ncsu.edu/unity/lockers/users/f/felder/public/Papers/CLCha
pter.pdf