Vous êtes sur la page 1sur 8

DESIGN OF AN ADAPTIVE E-LEARNING MODEL BASED ON

LEARNER’S PERSONALITY

Essaid El Bachari, El Hassan Abdelwahed, Mohamed El Adnani


Computer Systems Engineering Laboratory (LISI), Department of Enginneering Science,
Faculty of Science Semlalia, Cadi Ayyad University
B.P. 2390, Bd My Abdellah, 40000, Marrakesh
{elbachari, abdelwahed, md-eladnani }@ucam.ac.ma

ABSTRACT
Personalized e-learning implementation is recognized as among one of the most
interesting research areas in the distance Web-based education. Since the learning
style of each learner is different we must to fit e-learning to the different needs of
learners. This paper presents an Adaptive e-learning model based personality
learner’s. To recognize the learner’s personality, this system uses the Myers-
Briggs Type Indicator's (MBTI) personality dimensions. Secondly, it will propose
a Personalized Education System LearnFit Framework that suggests a learning
style matching with learner’s preference in online distance learning education.

Keywords: Adaptive Learning, E-learning, Learning Style, Teaching Strategy.

1 INTRODUCTION people prefer some kind of interacting with, taking


in, and processing stimuli or information.
Until now, it is extremely difficult for a teacher to
determine the optimal learning strategy for every 2 PREVIOUS WORKS
learner in a class. And even if a teacher is able to
determine all the strategies, it is even more difficult A lot of research works has been done about
to apply all multiple teaching strategies in a personality type, virtual learning system and learning
classroom. Today’s development of searching but it’s still very difficult to draw a definitive
technology provides learners a new way to break of conclusion on the relationship between them. For
the traditional educational models “one size fits all” example see [25], [11], [3], [21], [10], [8], [9] and
approach. It makes it possible to “customize down [24]. In [1], authors designed an interface for
to the individual” and hence for effective computer learners appropriate for the type of their
personalized and creative learning [30]. personality using MBTI test. Using learner’s
In response to individual needs, personalization in personality [23] and [15] proposed an expert system
education not only facilitates students to learn better for virtual Classmate Agent (VCA).
by using different strategies to create various Recent developments of the online learning are
learning experiences, but also teacher's designer’s related to Adaptive educational Hypermedia Systems
education needs in preparing or designing varied (AEHS). An AEHS aims at building a model of the
teaching or instructional packages. Each learner has goals, preferences and knowledge of each learner
a preference for a teaching style that allows him to and use this model throughout the interaction with
learn better. Some one likes to listen and talk, others the learner, in order to adapt learning content to the
prefer to analyze a text, or simply using a visual needs of that learner [6]. In this research, there are
medium. So to learn effectively, learners have to be four strategies path to improve the content suit for
aware of their preferences that make easy to manage individual learners. They are adaptive content,
their own way of learning. This information will adaptive navigation, adaptive presentation and
enable the learner to improve the effectiveness of its adaptive learning task.
approach to learning and to exploit its own resources. In this paper we will suggest new teaching
Jungian based psychologists add that people’s strategies on elearning context matching with
personality preferences influence the way they may learner’s personality using the Myers-Briggs Type
or may not want to become more actively involved in Indicator tools. In this purpose, we will apply our
their learning, as well as take responsibility for the approach that may include four strategy paths, to
self-direction and discipline. So we have to identify a implement a smart virtual learning which can give
person's individual learning style and then adapt learner and also teacher a new positive educational
instruction toward that person's strengths and experience.
preferences. It is commonly believed that most
3 THEORY BASE 4. Judging (J) or Perceiving (P)

The related studies about this research can be The Myers-Briggs Type Indicator reports a person’s
summarized under two sections; namely, the related preferences on four scales which, is given in Table 1.
studies about learning style and the related studies
about MBTI and personality learners. Table 1: The MBTI preferences and their definitions.

3.1 Learning styles Preferences Definition


The learning style (LS) can be defined as the Where a person prefer
way a person collects processes and organizes Extraversion or Introversion
to focus their attention
information. The way a person prefer
There are many models of learning styles Sensing or Intuition
to take in information
existing in literature. Individual learning styles differ, The way a person prefer
and these individual differences become even more Thinking or Feeling
to make decisions
important in the area of education. Learning style
How a person deal with
may be defined as “the attitudes and behaviours Judging or Perceiving
the external world
which determine an individual’s preferred way of
learning” [22]. Therefore, when an instructor's style
The various combinations of these preferences
matches a learner’s learning style; this affects the
result in a total of 16 personality types and are
learner’s experience and ability to do well. Until
typically denoted by four letters to represent a
today, a lot of research works has been done about
person’s tendencies on the four scales. For example,
learning styles and developed a good deal of learning
ENFP stands for Extroversion, Intuition, Feeling, and
style models but there does not seem to be any
Perceiving. This does not mean that a person
agreement of acceptance of any one theory [4]. Thus,
possesses only four preferences, but that the four
some tools are used to evaluate learner’s learning
preferences show a greater presence than their
style. There are many questionnaires that categorize
counterparts. The MBTI assessment can not only
people based on their learning preference. Indeed,
indicate the learner’s preferences, but also indicate,
MBTI is the well-known tool used for personality
how clear in expressing the preference for a
and learning style determination.
particular pole over its opposite.
For example, Fig. 1, typical report from
3.2 Evaluation of learning styles
Consulting Psychologists Press, E is showing a
Different tools are used to determine learners’
greater presence, on a clear level, over its opposite, I.
learning styles [2]. There are many questionnaires
N is showing a greater presence, on a moderate level,
that categorize each person according to their
over its opposite, S. F is showing a greater presence,
learning styles: Kolb questionnaire, honey and
on a clear level, over its opposite, T. Lastly, J is
Mumford questionnaire [20], GRSLSS questionnaire
showing a greater presence, on a moderate level,
[19] and [20]. Felder and Solman proposed a
over its opposite, P.
psychometric instrument, the index of learning style
Questionnaire ILSQ [16]. Among the different
proposals for modelling LS, we choose the MBTI
tools since it is the more powerful tools for personal
and team success.

3.3 The Myers-Briggs Type Indicator


The Myers-Briggs inventory is based on Carl
Figure 1: The strengths of MBTI type preferences.
Jung's theory of types, outlined in his 1921 work
Psychological Types [3], [12], [13] and [26]. Jung's
3.4 Dominant preferences
theory holds that human beings are either introverts
We all have an aspect of our personality which
or extraverts, and their behavior follows from these
dominates or governs us. It gives direction to the
inborn psychological types. He also believed that
personality and shapes our motives and goals. This is
people take in and process information different
called the Dominant Process. There is an Auxiliary
ways, based on their personality traits.
Process which should be the second in strength and
The Myers-Briggs evaluates personality type and
is the necessary assistant to the dominant. The
preference based on the four Jungian psychological
auxiliary takes care of the extraversion of the
types:
introvert and the introversion of the extravert. All
four of these processes are found in the middle two
1. Extraversion (E) or Introversion (I)
pairs of preferences, that is, the perceiving
2. Sensing (S) or Intuition (N)
preferences which are either Sensing or Intuition and
3. Thinking (T) or Feeling (F)
the judging preferences which are either Thinking or
Feeling. If a person has a Dominant judging process, benefit of the upcoming chapter's theory. Therefore,
his Auxiliary process will be one of the perceiving the teacher present the chapter's Theory or ideas, and
ones. Conversely, a person with a Dominant then applies it to the original application. Afterwards
perceiving process will have a judging preference for the teacher presents additional applications and has
his Auxiliary process [26]. Using these combinations the students apply the theory [5].
can reduce the number of sixteen personality types to
four. • Intuitive (N): people get information through
This is more manageable for planning teaching perception between relationships and results; usually
approach and monitoring learning engagement. use their conception to get information. They prefer
Extroverts (E) use their dominant preference mostly to know the theory before deciding that facts are
for the external world and introverts (I) use their important. Intuitive people will always ask “why”
dominant preference mostly for the inner world [3], before anything else. The discovery method, or the
[12], [13] and [26]. Table 2 illustrates clearly the why method, will appeal to intuitive students and
priorities and function for each of the 16 personality will teach sensing students how to uncover general
types. principles. To be motivated to learn, Intuitive people
prefer theory first, after the can attempt to analyze
Table 2: Priorities and direction of functions in each type. and solve the case or problem [5], [12].
Myers Briggs type Dominant preferences
3.5.2 Thinking-Feeling (T-F)
ISTJ, ISFJ, ESTP ,ESFP S This scale suggests the way a person makes
INFJ, INTJ, ENFP, ENTP N decisions.
ISFP, INFP, ESFJ, ENFJ F • Thinking (T): people that make their decisions
ISTP, INTP, ESTJ, ENTJ T based logic, analysis, and reason. Their
decisions are logical and impersonal [26]. Since
For example the dominant preference found in thinking students like clear course and topic
ISFP, INFP, ESFJ and ENFJ is sensing. For those objectives teacher have to present the chapter's
with ISTP, INTP, ESTJ or ENTJ type, the dominant Theory or ideas, and will use many examples
preference is thinking. For reasons of simplicity we that illustrate the concept [13], [12].
can define four MBTI classes or MBTI clusters. • Feeling (F): people that have emphasis on
)
S = {ISTJ, ISFJ, ESTP , ESFP} harmony and balance. They enjoy teamwork.
) Their judgments and decisions are based on
N = {INFJ, INTJ, ENFP, ENTP} personal value [26].
)
F = {ISFP, INFP, ESFJ, ENFJ}
) 3.6 Adaptive teaching strategies
T = {ISTP, INTP, ESTJ, ENTJ} We define the teaching strategies as the ways of
The proposed taxonomy of learner preference presenting instructional materials or conducting
consists on matching the different learning styles instructional activities. It will be designed in a way
with teaching strategies. Then it’s easy to suggest the that learner are encouraged to observe, analyse, look
suitable media as a channel for its representation, for a solution and discover knowledge by themselves
thus personalizing it to every learner. [5]. The main objective is to facilitate the learning
process.
3.5 Features of MBTI dimensions In the first step, learner encounters with system
Properties of each learner’s preference, questionnaire that finds out learner’s personality (for
pertaining to education and learning, were collated example ISEJ, ESTP, INTJ …), then the system sets
from the literatures [5], [12], [13], [14], [26] and learner in one of four MBTI taxonomy. Using learner
many others. traits based on Isabel Briggs Myers [3], [12], [13]
and [26], Table 3 below suggests an adaptive
3.5.1 Sensing-Intuitive (S-N) learning style scenario for each learner’s preference.
This scale suggests the way a person prefers to take
in information that can give idea on how he can learn.
• Sensing (S): people rely heavily on their five
senses to take in information. They like concrete
facts, organization, and structure. Applications
motivate sensing students to learn the material. To
be motivated to learn, Sensing people attempt to
analyze and solve the case or problem without the
Table 3: Learners personality and teaching style matched suggest.

MBTI Teaching strategies Suggest Distance educational delivery


cluster
• Instruction that allows them to hear and
TS1: it uses the Application-Theory-Application (ATA) approach. touch as well as see what they are
Teacher start by presenting an Application. The students attempt to learning
) analyze and solve the problem without the benefit of the upcoming • Hands-on labs, material that can be
S course's theory. Therefore, the teacher present the chapter's theory or handled
ideas, and then applies it to the original application. Afterwards the • Relevant films and other audio-visuals
teacher presents additional applications to make easy the learning • First hand experience that gives practice
process. for the skills and concepts to be learned

TS2: it uses the approach Theory-Application-Theory (TAT). • Assignments that put them on their own
Teacher start by presenting the chapter’s theory or idea before initiative
application related. The students attempt to analyze and solve the • Real choices in the ways they work out
) problem using the course's knowledge. The teacher can reuse the their assignments
N theory to facilitate the learning process. This approach is used for • Opportunities for self-instruction
the traditional educational model. individually or with a group
• Opportunities to be inventive and
original
TS3: it uses Theory- Example- Practical exercises Approach (TEP). • Logically constructed subject matter
Teacher start by presenting the chapter’s theory or idea before • Classrooms free from emotional
) examples related. The students attempt to analyze and solve the distractions
T practical exercises using the course's knowledge. Afterwards the • Interesting problems to analyze
teacher presents additional applications based logic and problem-
solving.
TS4: it uses the opposite teaching of TS3. In fact, it uses Practical • Having topics with a human angle
) exercises, Example then Theory (PET). Teacher use case studies or • Learning into harmonious small-group
F learning based practical exercises. Therefore, the teacher presents the work.
chapter's theory or ideas. • Collaborative work

4 PROPOSED FRAMEWORK MODEL As is shown in Fig. 2, the model includes three


main modules, each of them described below:
In this paper, a new Personalized Education • Module 1. Preference Engine: in first step, the
System LearnFit Framework is presented according system finds deals with detecting and storing the
to the learning model based on personality. This preferences in student model according to MBTI
module is displayed in Fig. 2. Our general purpose Tools. The Index of MBTI was added to the
framework may be viewed as being comprised of at registration form of Expert System for the future
least the following three elements: login.
a- Domain Model: Consist of concepts and the • Module 2. Adaptive Engine: according to learners
relations that exist between them. Typically the group, the system chooses a teaching style
domain model gives a domain expert’s view of matching with learner’s personality. This
domain. extension uses the decision unit to select an
b- Learner Model: Consists of relevant information adaptive course to improve the learning process.
about the user that is pertinent to the • Module 3. Revised strategy Engine: this model
personalisation of the learning style determines whether a given teaching style is
c- Pedagogical Model: includes two parts appropriate or not. For each individual student
• Adaptive Engine Model: Consists of a set of rules the system initializes the decision model
or triggers for describing the runtime behaviour generated from a set of rules that represents the
of the system as well as how the domain model matches between teaching style and the learner’s
relates to the user model to specify adaptation. personality. This model uses Bayesian Network
• Revised Strategy Model: Consists in determining (BN) and its behaviour is quite similar to a
whether a given resource is appropriate for a content-based recommender system1 [7].
specific learning style or not.
1
A recommender system tries to present to the user the
information items he/she is interested in. To do this the
Figure 3: Learner’s profile taxonomy

4.2 Domain Model


This model describes the structure of the
information content of the application. It consists of
Figure 2: System Architecture of LearnFit concepts and concept relationships. A concept is an
abstract representation of an information item from
The LearnFit project is an Add-On to the popular the application domain. The domain model of the
Moodle2 Learning Management System to provide system is based on the notion of learning goals that
adaptivity learning experience. The add-on is a web- the learner can select and study, and provides
based application having two tiers utilizing open learners with a plurality of learning activities and
source technologies: PHP, MySQL, XHTML, CSS, resources
and AJAX. Our framework has three meta-models: a In preview work, the authors in [27] suggest
learner model, a domain model and pedagogical three hierarchical levels of knowledge abstraction
model. Theses models will be described in the learning goals, concepts and educational material.
following sub-sections. We will add a fourth one that’s educational
pedagogical. This model is displayed in Fig. 4.
4.1 Learner Model
This component stores all user-related data, i.e.
the users’ profiles, including personal information,
preferences. It enables the system to deliver
customized instruction, on the basis of the individual
student’s, or the student group’s, learning style [29].
In our case, using MBTI questionnaire the
personality of the learner is recognized. The
questionnaire calculated and stored the preferences
in student model. It will only consider the four MBTI
) ) ) )
S
preferences of learners, which are , N , T and F .
The learner’s profile taxonomy is displayed in Fig. 3.

Figure 4: Teaching strategies suggestions for Myers-


Briggs Preferences.
user’s profile is compared to some reference
characteristics. These characteristics may be from the A learning goal corresponds to a topic of the
information item (the content-based approach) or the domain knowledge, which can be recognized and
user’s social environment (the collaborative filtering selected even by a novice learner. Each goal is
approach). associated with a subset of concepts of the domain
2 Moodle has been an open source Learning managing knowledge, which formulates a conceptual structure
system designed to help educators to have a platform that represents all the concepts of a goal and their
where they could create online courses. Moodle is been relationships [27].
developed all the time by several parties who create new Teaching strategies hold a one-to-one
qualities and improve existing ones.
relationship with the learning styles. There can be
one teaching strategy that accommodates one
learning style. Each concept is associated with a
teaching strategy according to learner’s preferences
TSD = {TS 1 , TS 2 , TS 3 , TS 4 } (1)
Where TSD denotes the set of all teaching strategies
used in our framework.
Each teaching style TSi can be also associated
with appropriate learning objects (LOi):

{ i i i i
LO i = LO1 , LO2 , LO3 , LO4 ,..., LOn
i
} (2)

Where LOi denotes the set of all learning objects


related to each teaching strategy TSi. Figure 5: Adaptive taxonomy: LS dimensions and
Otherwise a teaching scenario can be also TS relationships
represented by a tree of learning object where
sections constitute the components (Child) of 4.3.2 Revisited strategy engine
learning scenarios, each section can be composed of The revisited strategy engine helps to determine
subsection that can also be composed of more whether a given teaching strategy is appropriate for a
specific learning objects like pedagogical resources specific learning style or not. This module uses a
and/or pedagogical activities. Dynamic Bayesian Network Classifiers DBN 3 [7],
[18] and [28] to classify a teaching strategy as
4.3 Pedagogical Model “appropriate” or “not appropriate” for the learner.
Teaching strategies (TS) are the elements given To define the DBN’s parameters we set the a priori
to the students by the teachers to facilitate a deeper distribution of the nodes representing the LS
understanding of the information. The emphasis according to the score obtained by learner in the
relies on the design, programming, elaboration and MBTI test (related module 1).
accomplishment of the learning content. Teaching
strategies must be designed in a way that students are
encouraged to observe, analyze, express an opinion,
create a hypothesis, look for a solution and discover
knowledge by themselves. Didactic teaching strategy
for example refers to an organized and systematized
sequence of activities and resources that teachers use
while teaching [18]. The main objective is to
facilitate the students´ learning. Our pedagogical
model has two main intelligent axes: adaptive
strategy engine and revisited strategy engine.

4.3.1 Adaptive strategy engine


The system is designed to offer a best teaching Figure 6: Bayesian networks representing the
environment matching with the learner’s profile. The decision model
system sets learner in one of four kinds of
independent group given in Fig. 5. This classification For a given concept, the system offers a suitable
is obtained based on preference engine which teaching style (Selected Ti), the value of each node
detecting and storing the preferences in student can be changed dynamically when the concept’s
model according to MBTI Tools. If the learner be score is acceptable. For example, Fig. 6 shows that for a
)
placed in the independent group, the education a person with S profile, the system offers a selected teaching
learning process will be started otherwise system style selected T1. Thereafter, if the student's grade is less
selects a suitable teaching strategy according to than 60%, the system presents another teaching style
learner’s personality type. for example selected T2, and so until the score will
be more acceptable. In that time, the value of the
learner’s profile can be adjusted.

3 A DBN is a model to describe a system that is


dynamically changing or evolving over time. This model
enables uses to monitor and update the system as time
proceeds.
Regarding the conditional probability tables storing the preferences in student model according to
(CPTs) that represent the relationships between the MBTI Tools. Then, according to learners group, the
dimensions of the LS and the TS, we estimated these system chooses a teaching style matching with
CPTs taking into account the matching tables defined learner’s personality. This extension deals with the
by the expert. The detail will be published in the next decision unit to select the best adaptive course.
work. Lastly, the system determines whether a given
teaching style is appropriate or not using Bayesian
5 SYSTEM EVALUATION network (BN).

To evaluate the performance and the impact of We are very encouraged by the system’s ability
the LearnFit framework to improve learning process, and concentrating on evaluating the effectiveness
we plan to compare the learning outcome of and impact of Learnfit’s to facilitate a positive
LearnFit’s adaptive teaching strategy with a classical educational experiences. The result of this
non adaptive strategy. The difference will be experimentation will be published in the next work.
measured with respect to the following two main
criteria: 7 REFERENCES
• Does the adaptive teaching strategy contribute to
a successful learning experience [1] E. Abrahamian, J. Weinberg, M. Grady and C.
• Does the impact of personality help educator to M. Stanton: The Effect of Personality-Aware
making an effective group in collaborative Computer-Human Interfaces on Learning, JUCP,
environment 10:27–37 (2004)
We have set up experimentation for this semester [2] T. Anderson and F. Elloumi: Theory and
to compare our system with a classical one, by practice of online learning (2003)
measuring the student understanding after learning [3] C. Bishop-Clark, D. Wheeler: The Myers-Briggs
process. Future work will involve a more exhaustive Personality Type and its relationship to computer
evaluation, both with simulated and real learners. programming. JRCE. 26:358-370 (1994)
This evaluation must account also for changes in [4] C. Bruen and O. Conlan: Adaptive ICT Support
learners’ preferences. Once the model has been for learning Styles- A Development Framework
validated, it will be used in a decision model aimed for re-useable learning resources for different
at determining the more learning objects (LO) for learning styles & requirements, AITTE (2002)
each learner. [5] H. J. Brightman : Mentoring Faculty to Improve
Teaching and Student Learning Decision
6 CONCLUSION Sciences. JIE(2005). Retrieved 2 May 2010,
from http:// gsu.edu/~dschjb/wwwmbti.html
Utilizing traditional web-based system to support [6] P. Brusilovsky: Methods and Techniques of
learning is one useful approach in education. The Adaptive Hypermedia. User Modeling and User-
problem is that, during on-line learning, a massive Adapted Interaction, KAP 87-129 (1996)
amount of content presented will make learners feel [7] C. Carmona, G. Castillo and E. Millán:
frustrated and dissatisfied. The problem of Discovering Student Preferences in E-learning,
personalization in eLearning remains the focus of EC-TEL07, pp.33-42 (2007)
attention of many researchers nowadays. Various [8] S. Chaffar, G. Cepeda and C. Frasson:
attempts have been undertaken but only some of the Predicting the Learner’s Emotional Reaction
solutions are practically useful for teaching. Towards the Tutor’s Intervention. 7th IEEE Inter.
Sophisticated web-based Adaptive Hypermedia Conf., Japan, pp. 639–641 (2007)
systems as well as Intelligent Tutorial systems are [9] S. Chaffar and C. Frasson: Inducing Optimal
often oriented on one type of task, for instance Emotional State for Learning in Intelligent
quizzes or assessments, and therefore cannot be used Tutoring Systems. Comp. Sci., pp. 45–54 (2004)
for other purposes [6]. In this paper another [10] P. Chalfoun, S. Chaffar and C. Frasson:
approach has been presented. In this model some Predicting the Emotional Reaction of the Learner
modules for personality recognition and selecting an with a Machine Learning Technique. Workshop
appropriate teaching strategy for learner’s on Motivational and Affective Issues in ITS.
personality, we considered to develop interaction Taiwan. (2006)
with the learner. The first result shows that placing [11] M. Crosby, J. Stelovsky: From Multimedia
the learner beside an appropriate learning style Instruction to Multimedia Evaluation. 4:147-162.
matching with learner’s preference in online distance (1995)
learning education leads to improvement and makes [12] J. K. DiTiberio: Education, learning styles, and
the virtual learning environment more enjoyable. cognitive styles. In MBTI applications: A decade
LearnFit uses three modules for strategy adaptation. of research on the Myers-Briggs Type Indicator,
First, the system finds deals with detecting and ed. A. L. Hammer. Palo Alto: CPP Press (1996)
[13] J. K DiTiberio: Uses of type in education. In [29] D. Sampson, C. Karagiannidis and F.
MBTI Manual: A guide to the development and Cardinali: An architecture for web-based e-
use of the Myers-Briggs Type indicator, eds. I. B. learning promoting re-utilizable adaptive
Myers, M. H. McC., and N. Q. CPP (1998) educational e-content. JETS. 5 27-37(2002)
[14] J.K. DiTiberio and A.L. Hammer: Introduction [30] T.I. Wang, K.T.Wang and Y.M. Huang: Using a
to type in college. Palo Alto, CPP (1993) style-based ant colony system for adaptive
[15] S. Fatahil, M. Kazemifard1, N and Ghasem- learning. ESWA, 2449-2464 (2008).
Aghaee1.: Design and Implementation of an
ELearning Model by Considering Learner's
Personality and Emotions. AEECS 423-434
(2009)
[16] R.M. Felder, L.K. Silverman.: Learning and
Teaching Styles in Engineering Education, EE,
78 (7), 674-681 (1988)
[17] R.M. Felder and B.A. Soloman: Index of
Learning Style Questionnaire. Online version
(1997). Retrieved 2 May 2010, from
http://www.engr.ncsu.edu/learningstyles/ilsweb.
html
[18] A.L. Franzoni, A.L. and S. Assar: Student
Learning styles Adaptation Method on Teaching
strategies and Electronic Media, Edu. Tech. Soc.
12 (4) 15-29 (2009)
[19] H.G Lang, H.G., M.S Stinson, M.S., F.
Kavanagh, Y. Liu, M. Basile: Learning style of
deaf college students and Deaf Education, Vol.,
4, pp. 16-27 (1999)
[20] K. Logan, K., P. Thomas: Learning style in
Distance Education Students Learning to
Program, Proc. Work. Psy. Prog. Int. group,
Brunel University, pp. 29-44 (2002)
[21] J. Gurka, W. Citrin: Testing Effectiveness of
Algorithm Animation. In PSVL. 360-367. (1996)
[22] P. Honey and A. Mumford.: The Manual of
Learning Styles. Maidenhead, Berkshire (1992)
[23] H. Maldonado, J.R. Lee, S. Brave, C. Nass, H.
Nakajima, R. Yamada, K. Iwamura and Y.
Morishima: We Learn Better Together:
Enhancing e-Learning with Emotional
Characters. In Proceedings, Comp. Supp.
Collab.Learning, Taipei, Taiwan (2005)
[24] B.F. Marin, A. Hunger and S. Werner:
Corroborating Emotion Theory with Role
Theory and Agent Technology. A Framework
for Designing Emotional Agents as Tutoring
Entities, J. Net. (1), 29–40 (2006)
[25] K. F. Matta, G. M. Kern: Interactive videodisk
instruction: the influence of personality on
learning. IJMMS, 35: 541-552 (1991)
[26] I. Myers-Briggs.: Introduction to Type. 5th ed.,
Palo Alto, CA: CPP (1993)
[27] K.A.Papanikolaou, M. Grigoriadou, H.
Kornilakis and G.D. Magoulas: Personalizing
the interaction in a Web-based educational
hypermedia system: the case of INSPIRE.
UMUAIU. 13 (3) 213-267 (2003)
[28] M. Saham: Learning Limited Dependence
Bayesian Classifiers., KDD-96 AAAI Press 335-
338 (1996)

Vous aimerez peut-être aussi