Gamification of e-learning based on learning styles –design model and implementation

Abstract

In recent years, there has been an increasing interest in applying game-related principles in Technology Enhanced Learning (TEL). Implementation of game elements in non-game contexts i.e. the gamification, has tremendous potential in the education space and have emerged as a powerful application for improving the learning experience. In addition, current researches are investigating cross-references between personalized learning, learning analytics and gamification in order to create an optimal framework for TEL. In this manner a term Gamification 3.0 raises, referring to delivery of personalized and richer, contextually relevant gamified experience. In this paper, we propose the Learning Style Gamification Model (LSGM) for higher education based on students' learning styles and their actions and behavior in e-learning environments. Case study approach with the qualitative and quantitative analysis is used to identify which game elements could provide a positive influence on students' learning for specific learning style.

Full text

Gamification of e-learning based on learning styles –design model and
implementation
Nadja Zaric
RWTH Aachen University, Germany
zaric@informatik.rwth-aachen.de
Snezana Scepanovic
University Mediterranean, Podgorica, Montenegro
snezana.scepanovic@unimediteran.net
Abstract: In recent years, there has been an increasing interest in applying game-related principles
in Technology Enhanced Learning (TEL). Implementation of game elements in non-game contexts
i.e. the gamification, has tremendous potential in the education space and have emerged as a
powerful application for improving the learning experience. In addition, current researches are
investigating cross-references between personalized learning, learning analytics and gamification in
order to create an optimal framework for TEL. In this manner a term Gamification 3.0 raises,
referring to delivery of personalized and richer, contextually relevant gamified experience. In this
paper, we propose the Learning Style Gamification Model (LSGM) for higher education based on
students’ learning styles and their actions and behavior in e-learning environments. Case study
approach with the qualitative and quantitative analysis is used to identify which game elements
could provide a positive influence on students’ learning for specific learning style.
Introduction
Games have been known to engage and interest people since its inception. The advent of games and their
evolution through the ages have only increased its audience over time. Its popularity, among the masses, invoked
market experts to explore the possibility to implement game principles in non-game contexts in order to grab the
attention of their consumers (Flatla et al., 2011). In this manner, a term gamification rises, referring to the concept of
using game elements in non-game contexts (Education & eLearning 2.0, 2018). The main goal of using game
elements in contexts other than play is to boost motivation, interest and to encourage users to engage with the
environment in which elements are implemented. By using game-like techniques such as scoreboards, awards or
personalized fast feedback gamification can make people feel a sense of ownership and purpose when engaging with
the tasks or product (Muntean, 2011). When it comes to education, because of its implementation possibilities with
flexible resources and protocols, gamification is mostly related to TEL and serves mainly as a tool for increasing
motivation and engagement among students (Martens& Mueller,2016). Research studies, confirmed that, if
implemented correctly, gamification can make a positive impact on students’ motivation and participation (Barata et
al., 2013; Fan,Xiao&Su, 2015; Dicheva et al., 2015) . Still, the implementation is mainly presented on individual
cases and its success differs from study to study. Depending on the environment, participants, learning materials or
learning goals, gamification “rules” may change as well as its effect. The main challenge in gamifying e-learning
lies in students' diversity i.e. in a fact that students learn, behave and act in different ways (Çağlar& Kocadere,2016).
In order for gamification to be successful, those diversities need to be considered i.e. the fact that students have
different learning styles (LS) and behavior implies that they’ll have different reactions and attitudes toward
gamification. This new approach to gamification, where different users’ characteristics are considered before and
during the incorporation of game elements is a so-called Gamification 3.0 (Cognizant, 2017)
This paper presents the results of implementation and limitations of proposed LSGM model, which aims to
create a personalized gamified e-learning environment in which learning materials are adjusted to student' LS and
their behavior in the e-learning environment.
Learning styles
LS represent psychological and cognitive behaviors that serve as indicators for measuring the way learners
learn and interact with the learning environment (Pashler et al., 2008). LS are derived from learning theories, mostly
from behaviorism, constructivism and the cognitive theory (Cannarelli, Kahn& Schneider, 2016). Combining the
Preview version of this paper. Content and pagination may change prior to final publication.
E-Learn 2018 - Las Vegas, NV, United States, October 15-18, 2018

principles of those theories LS models are created that suggest how and in which group learner can be classified. In
recent years, numerous literature claims that matching LS with an instructional medium for an individual positively
impacts their performance and helps them to attain their learning goals (Cassidy, 2004; Wang & Mendori, 2015).
However, some research papers claim that LS must be used carefully and not as a panacea to fix learning difficulties
(Pashler et al., 2008). Pashler et al. (2008) provide a strict methodology on how the research on LS impact should
be conducted and measured in order for its results to be scientifically acceptable “we think the primary focus should
be on identifying and introducing the experiences, activities, and challenges that enhance everybody’s learning” and
that “learning styles may be important but cannot be seen as a standalone solution for overcoming learning
difficulties” (p.117). Learning style-based approach is gaining more and more popularity in both formal and
informal education, but still there are a lot of issues that need to be considered like age of student, culture influences
and availability of technological advances (eg, personal digital assistants, mobile phones, the World Wide Web,
wireless Internet) that make younger generations of students more comfortable with e-learning)
Felder-Silverman Learning Style Model
In this study Felder-Silverman Learning Style Model (FSLSM) is chosen as a tool for identifying student
LS. Important reasons for selecting this model are described in previous authors work "FSLSM combines major
learning style models such as the ones by Kolb , Pask , and Myers-Briggs . Furthermore, FSLSM is one of the most
often used LS model in technology enhance learning"(Scepanovic & Debevc, 2012, p. 5061) . Felder and Silverman
distinguish four basic categories of LS: active-reflective, verbal-visual, sensor-intuitive and sequential-global. The
LS categorization (Felder & Silverman, 1998) present students’ preferences on four dimensions: a) Perception -
how they prefer to perceive or take in information, b) Input - how they prefer information to be presented, c)
Processing - how they prefer to process information, d) Understanding - how they prefer to organize and progress
toward understanding information. For identifying preferences on abovementioned dimensions, Felder (2010)
designed a web-based questionnaire Index of LS (ILS). Based on scores, students can have mild, moderate or strong
preferences to one of the dimensions. Mild and moderate preferences mean that a student feels more or less
comfortable in both styles of dimensions, while those with strong preferences may struggle and not manage to
realize their potential in the learning environment which is not adapted to their prefer LS.
Related work
Since the inception of LS theories, their impact in both traditional classrooms and e-learning is being
studied. For example, Fan, Xiao &Su (2015) used Kolb's LS model as a research instrument to study the learner
preferences and develop game-based mobile learning application which teaches the blood circulation system to
Junior high school students. The authors suggest curriculum design and a combination of teaching strategies for
mobile game-based learning activities adapted to learners with different LS. Sousa et al. (2016) designed an
experiment to measure the impact of LS in gamified learning. It evaluates Felder-Silverman LS theory using a lean
knowledge dissemination model. The study reports that the overall feedback provided was a positive and significant
portion of the students (92%) considered the session was engaging and motivating. Buckley& Doyle (2017) used
Felder- Silverman model and gamified learning tool in the study to evaluate whether an individual's personality
traits and LS affect their experience of gamification. The authors conclude that gamification needs to be integrated
into learning activities carefully; gamification should not be introduced as a standalone activity as it would
negatively affect learner with other LS.
From the above mention literature, we derive that only the studies based on Felder- Silverman model fits
into the research domain of analyzing the effect of LS in gamification. All of the studies discussed in this section
support incorporating gamification and learning style theories in creating curriculum and teaching strategies.
However, practical experiences from case study examples are mainly focused on developing educational games for
learning a specific topic, without taking in mind gamification of the curriculum. Evidently, there is a gap in the
research area of gamification based on the LS of students' and further research is needed, as so, more empirical
studies must be conducted to conclude the efficacy of LS in gamification with confidence.
Learning Style Gamification Model (LSGM)
Our model is based on gamification 3.0 concept and proposes personalization of e-learning courses with
game elements that are suitable to the learning style of student defined by the FSLSM. Conceptual design of model
Preview version of this paper. Content and pagination may change prior to final publication.
E-Learn 2018 - Las Vegas, NV, United States, October 15-18, 2018

and methodology for the development of gamified e-learning courses based on LS are described in author’s previous
publication (Zaric et al., 2017). LSG model proposes an architecture of virtual learning environment (VLE) for
creating gamified e-learning courses in which LS are base for personalization of course content. The first step in
creating the design of gamified e-learning is defining student profile with the user model. The user model combines
the ILS questionnaire and data analysis of prior knowledge (grades, assessment, pre-knowledge tests, behavior in
VLE) for identifying students LS and determine student profile. As the proposed model is focused on introducing
game elements based on students’ learning style, the second step is the design and development of lectures and
assignments based on the LS of students while third step deals with incorporation of gamification in the course.
Implementation and testing of model
In order to test the proposed model, case study research was conducted during summer semester 2018.
with the first year students of the Faculty of Information Technology, University "Mediterranean" Podgorica. A
mandatory course titled "Introduction to Web Technologies" was selected for a pilot case study. The course covers
theoretical foundations of web technologies as well as the practical application of HTML and CSS technologies. The
main goal of the pilot study was to test the following hypothesis:
H1: Students who participate in gamified e-learning course will achieve better learning outcomes
comparing to participants from non-gamified e-learning course.
H2: Game elements will have a different influence on different LS.
H3: Gamification will have a different impact on overall satisfaction and motivation among different LS.
During the pilot study, we followed the recommendations from Pashler et al. (2008) where authors
suggested that results of the experiment are acceptable only if two same LS achieve different results while treated
with different learning method. As so, students from the Group A attended a gamified e-learning course, while those
from Group B attended an e-learning course without game elements. All the knowledge assessment test were the
same for both groups and without game elements, so the results could be compared properly. At the beginning of the
e-learning course, students’ were given two tests – ILS questionnaire and pre-knowledge test (in order to define their
prior knowledge related to course’ matter). Based on the results, participants were divided into two groups:
experimental (Group A) and control (Group B) with 39 members each. Both groups had relatively the same
distribution of LS and approximately the same average score on the pre-knowledge test (7.73/10.00 in the
experimental group and 7.51/10.00 in the control group). For both groups of students, the course syllabus was the
same but the design of e-learning materials was different. Control group of students used e-learning materials in
form of pdf, video lectures, and online code examples for exercise. Other , experimental group used the same e-
learning materials but with game elements.
E-learning course design
The gamified e-learning course "Introduction to Web Technologies" was created in Moodle VLE, a
learning management system (LMS) already used at Faculty for Information Technologies, University
“Mediterranean”. In this way, we provided students with the familiar user interface for e-learning. The course was
designed as a linear e-learning course organized in 4 learning units – each learning unit was defined as level.
Learning unit can support its own learning objectives without relying upon the content of the other e-learning
modules or lessons that come before or after it. In our pilot study student couldn’t start a new learning unit (level)
until he successfully the previous lesson. Every learning unit consists of lectures, learning activities and practical
assignment or a test . For each type of learning style, corresponding learning resource and/or activity was created
according to the FSLSM recommendation. For gamification of e-learning course, we have used the most popular
game elements such as:
Levels and experience points. For each level (learning unit) we have defined rules for gaining experience
points (XP’s) needed to achieve the next level. Student receives a certain number of experience points (XP’s) for
each successfully completed learning activity. In pilot study students had to reach a certain number of XP’s to open
a new learning unit
Awards (Badges). Badges were defined based on different criteria such as: successfully completed
assignment, accumulation of the certain amount of experience points, participation in discussion and similar student
activities . Every badge has the name, detailed description and criteria for the award.
Preview version of this paper. Content and pagination may change prior to final publication.
E-Learn 2018 - Las Vegas, NV, United States, October 15-18, 2018

Progress track. The progress track shows all the activities that a student needs to complete in course. The
progress is shown as a three-color bar: green color indicates the successfully completed activities, red color indicates
un-successfully completed activities, while blue color shows activities that have not been started yet.
Leaderboards: Leaderboard scale was shown inside the student’s personal profile, showing names and
results for each peer of its group. Results were calculated based on XP’s, where a student with highest XPs amount
was ranked as first on the board.
Mystery. For the implementation of the game element “mystery”, password protected test is used. The
password was hidden in a glossary of the e-learning course. In order to find the password for a test, students need to
use game instructions for searching and combining specific e-learning course materials. During the game of finding
password students learn to combine knowledge from different learning units.
Analysis and evaluation
For the purpose of analyzing hypothesis (H1, H2, and H3), data from three sources were collected for each
student in control and experimental group: scores from the ILS questionnaire, results of the pre-knowledge test,
results of e-learning activities and self –examination/exercise knowledge tests. Pearson’s Chi-Square test
(Lancaster&Seneta, 1986) is used to find correlations between variables, in order to measure the impact of
gamification on LS, learning outcomes and general satisfaction and motivation of students for e-learning.
Gamification and learning outcomes: We have conducted a comparative analysis of the students’
achievements within e-learning course (learning activities and tests) without taking LS into account. The
experimental group achieved an average score of 8.5 / 10.00, while the control group for the same, but non-gamified
tasks achieved 7.3 / 10.00. Further, the percentage of students who failed in Group A is 4.8%, while in Group B, it is
slightly higher: 5.7%. The results show that the gamified materials had a positive impact on the learning outcomes
of students, that is, the learning outcomes in a gamified environment were better than in non-gamified e-
environment. Moreover, results showed that group A achieved the best results (9.0 / 10.0) in the second learning
unit, while group B achieved the worst results in it (6.3 / 10.0). By weighting the variable group with the results of
individual tests, statistical significance (p = 0,019) has been noted. Namely, in the case of assessment-test in the
learning unit II, the success depended on the group to which student belonged, i.e. the achieved result depended on
whether the learning unit was gamified or not. In particular, learning material in unit II was organized through
levels, where a student had to successfully finish one learning chapter before he/she proceeds to next. Based on
results we can see that this linear kind of material organization had directly influenced the outcomes of mastering
the material. In the first learning unit, for the group A, XP’s were given for visiting external, additional information
on the subject of matter. In the control group, additional material was also given but without any rewards. At the end
of the lesson, a self-assessment test was given with questions related to the additional material. This test was
optional and its results were not taken into account for the final grade calculation. In this test, group A achieved an
average score of 8.2 / 10.00, while the control group achieved 6.7/10.00. Those results imply that XPs points as a
reward mechanic can encourage students to devote more time and assign more force while dealing with the new
material.
Gamification and learning styles: In this step we have tested H2 in order to find out the influence of game
elements to different LS. Results of experiment showed that the best overall results (final grade > 9.0) on the course
level, in the Group A were achieved by active, sequential, intuitive and visual learners, opposite to Group B, where
the best results achieved reflective, sensitive and global learners. By weighting the variable learning style and final
grade in Group A, statistical significance was obtained. Active and sequential learners achieved significantly better
results in gamified environment comparing to ones from the control group (p = 0.019 in case of Active/Reflective
and p = 0.018 in case of Sequential/Global). In other words, active and sequential learners in group A (gamified
environment) achieved significantly better results from their same-learning-style peers in group B. Opposite is the
case for the global and reflective learners. In order to provide a better insight into the correlations between game
elements, LS and learning success, analyses of the individual activities, as well as assessment-test, were necessary.
Results showed that, in a group A, badges encouraged reflective students (who were awarded for participating in
forum discussions) to take a part in this collaborative activities twice as much, comparing to their peers from group
B. For "HTML quiz", a somewhat higher failure level was observed in case of global and intuitive learners
comparing to other LS , 37% of students had less than 6/10 score. The quiz was conducted after a lesson organized
in levels, which allows presumption that learning materials presented in form of levels have a negative impact on
learning outcomes within global and intuitive learners. Further, a negative impact was also recognized in the case of
the mystery element. The results showed that 73% of sensing learners gave up after less than 5minutes of searching
Preview version of this paper. Content and pagination may change prior to final publication.
E-Learn 2018 - Las Vegas, NV, United States, October 15-18, 2018

for a password. Based on the analysis we have created a correlation between game elements and LS which are
presented in Table 1.
Active Reflectiv
e Visual Verba
l Sequential Global LS Sensing LS Intuitive LS
Levels - x x - x o x o
Feedback x x x x - - - x
Mystery - - - x x x o x
Time limit x x - - x - o -
Table 1. Positive (x), negative (o), no significant (-) impact of game elements to different LS
The influence of gamification on students’ overall motivation and satisfaction: For the purpose of
examining students’ attitudes and opinions towards the realized gamified course, students from the experimental
group were asked to give their opinion on pre-defined statements using Liker’s scale (Fieldboom, 2010). Results
show that the general attitude of the students toward the gamification of learning content is positive. Also, for each
element of the game, the influence on motivation, engagement and general satisfaction is assessed mainly as
positive. In evaluating the individual elements of the game the results showed that the element "mystery" did not
have a negative impact on learning outcomes, but it did on the overall satisfaction and interest of students. In the
case of badges, ranking lists, and experience points, the impact on motivation was assessed as positive or mainly
positive (80-95%). However, no statistically significant dependence of responses on LS was observed. Regarding to
general satisfaction and attitude in relation to the gamification, 78.9% of students rated with mark 4 or 5 the
statement "Game elements have increased my interest in learning". Further, 86.3% of students would like to include
elements of the game in all the courses they attend, while 72% of them think that the course with game elements suit
their preferred way of learning.
Conclusion and future work
In this research, we have presented reference model for creating gamified e-learning courses according to
LS of students that can provide instructors with guidelines how to define LS of students and which game elements to
choose for different LS. The conducted pilot study showed that gamification can affect students’ behavior in VLE
and overall results achieved with different learning activities. In terms of LS, this research found a negligible
number of examples of how the impact of the game element can differ from one learning style to another. We found
that game element badges can encourage students to read additional materials which will eventually lead to gaining
more knowledge. Further, creating a lesson in form of levels can enforce students to learn step by step, without
skipping any part of the learning unit. Assigning experience points for additional tasks can encourage students to
take more learning tasks or participate in collaborative actions like forums and chat. Game element mystery made a
positive impact on intuitive students, but reduced motivation among sensing students. On the other side, badges
encouraged reflective students to take part in the discussion even though they are not likely to do that. Sharing
knowledge among peers is important in every learning process, and we recommend award systems to be
implemented to encourage those kinds of activities. Regardless of the system of learning, lifelong learning or formal
education, gamification presents opportunities to perform learning in innovative and personalized ways. The results
of research presented in this paper contribute to the fact that the use of game elements in developing a model for e-
learning and teaching can significantly contribute to the success of learning if they comply with the LS of students.
The presented results are encouraging for future research towards the personalization of gamification. Special
attention needs to be put on learning analytics, as a way to measure the effects of gamification in a more detailed
manner. In next step, we will develop tools for monitoring the behavior, activities, and interaction of students within
the e-learning platform in relation to their LS. Tools for visualization will provide better analyzes of qualitative data
with detailed recommendations on instructional design in real time.
References
Preview version of this paper. Content and pagination may change prior to final publication.
E-Learn 2018 - Las Vegas, NV, United States, October 15-18, 2018

Barata, G., Gama, S., Jorge, J., & Gonçalves, D. (2013). Improving Participation and Learning with Gamification.
In Proceedings of the First International Conference on Gameful Design, Research, and Applications, New York,
NY, USA: ACM, pp. 10–17.DOI: https://doi.org/10.1145/2583008.2583010
Boller, S. (2013). Learning Game Design: Game Elements. Retrieved February 10, 2018, from
http://www.theknowledgeguru.com/learning-game-design-game-elements/
Buckley, P., Doyle, E. (2017). Individualising gamification: An investigation of the impact of learning styles and
personality traits on the efficacy of gamification using a prediction market. Computers & Education, 106, pp.43-55.
DOI: 10.1016/j.compedu.2016.11.009
Çağlar, S., & Kocadere, S. A. (2016). The possibility of motivating different type of players in gamified learning
environments. In Proceedings of EDULEARN16 , pp.1987–1994.
Cassidy, S. (2004). Learning Styles. An overview of theories, models, and measures. Educational Psychology, 24(4)
pp. 419-444
De Vita, G.D. (2001). Learning styles, culture and inclusive instruction in the multicultural classroom: a business
and management perspective. Innovations of Education and Teaching International, 38, pp.165–174.DOI:
10.1080/14703290110035437
Dicheva, D., & Dichev, C. (2015). Gamification in Education: Where Are We in 2015? In Proceedings of E-Learn:
World Conference on E-Learning in Corporate, Government, Healthcare, and Higher Education 2015, Chesapeake,
VA: Association for the Advancement of Computing in Education (AACE), pp. 1445-1454.
Education & eLearning 2.0: A Brief History of Gamification: Part I - The Origin. (n.d.). Retrieved May 8, 2018,
from http://edulearning2.blogspot.com/2014/03/a-brief-history-of-gamification-part-i.html
Fan, K., Xiao, P., Su, C. (2015). The Effects of Learning Styles and Meaningful Learning on the Learning
Achievement of Gamification Health Education Curriculum. Eurasia Journal of Mathematics, Science and
Technology Education, 11(5), pp.1211-1229. DOI: 10.12973/eurasia.2015.1413a
Felder, R., Silverman, L. (1998). Learning and Teaching Styles in Engineering Education. Engineering Education,
78(7), pp. 674-681
Fieldboom Blog (2010). The Likert Scale Explained — With Examples & Sample Questions (2010), Retrieved May
10, 2017, from https://www.fieldboom.com/blog/likert-scale/, last accessed 2018/01/14
Flatla, D. R., Gutwin, C., Nacke, L. E., Bateman, S., Mandryk, R. L. (2011). Calibration games: making calibration
tasks enjoyable by adding motivating game elements. In Proceedings of the 24th annual ACM symposium on User
interface software and technology (UIST '11). ACM, New York, NY, USA, pp. 403-412. DOI:
https://doi.org/10.1145/2047196.2047248
Lancaster, H. O., Seneta, E.(1986). Chi
‐
Square Distribution, John Wiley & Sons, Ltd.
Martens, A., & Mueller, W. (2016). Gamification - A Structured Analysis. In J. M. Spector, C. C. Tsai, D. G.
Sampson, Kinshuk, R. Huang, N. S. Chen, & P. Resta (Eds.), IEEE 16th International Conference on Advanced
Learning Technologies (ICALT), pp. 138–142.
Muntean, C. (2011). Raising engagement in e-learning through gamification. In Proc. 6th International Conference
on Virtual Learning ICVL, pp. 323-329.
Preview version of this paper. Content and pagination may change prior to final publication.
E-Learn 2018 - Las Vegas, NV, United States, October 15-18, 2018

Pashler, H., McDaniel, M., Rohrer, D., Bjork, R. (2008). Learning Styles: Concepts and Evidence. Psychological
Science in the Public Interest: A Journal of the American Psychological Society, 9(3), pp.105–119. DOI:
10.1111/j.1539-6053.2009.01038.x
Sousa, R.M., Stadnicka, D., Dinis-Carvalho, J., Ratnayake, R. M. C., Isoherranen, V. (2016). Gamification based
lean knowledge dissemination: A case study. Proceedings of IEEE International Conference on Industrial
Engineering and Engineering Management (IEEM 2016), pp. 164-168.
Zaric, N., Scepanovic, S., Vujicic, T., Ljucovic, J., Davcev, D. (2017). The Model for Gamification of E-learning in
Higher Education Based on Learning Styles. In Proc. Of ICT Innovations 2017 , Springer, Cham. pp. 265–273.
DOI: 10.1007/978-3-319-67597-8_25
Preview version of this paper. Content and pagination may change prior to final publication.
E-Learn 2018 - Las Vegas, NV, United States, October 15-18, 2018