Yousuf M. Islam, PhD
Director, Research & Advisory Services, BRAC University, Bangladesh
Mawdudur Rahman, PhD
Director of Online Programs and Professor of Accounting, Suffolk University
Zillur Rahman
Lecturer, Dept of CSE, BRAC University
Manzur Ashraf
Lecturer, Dept of CSE, BRAC University
Keywords: Distance learning, e-learning, Inf
ormation system, Application of Mobile Telephone technology.
Abstract: This paper presents the methodology, results and effectiveness in the develop
ment of Mobile telephone-
based (Short Message Service-based) distance learning. The proposed novel real time interactive distance
learning approach is about the application of information technology to education, was setup, delivered and
evaluated using a real-life environment. Statistical analysis of the achieved results of the learners confirmed
this SMS-mobile based learning being as effective as direct face-to-face learning.
The term Distance Learning (DL) means different
things to different people. As the meaning of the
term varies so also the delivery techniques used by
different users and institutions. The California
Distance Learning Project defined distance learning
as: "Distance Learning (DL) is an instructional
delivery system which connects learners with
educational resources. DL provides educational
access to learners not enrolled in educational
institutions and can augment the learning
opportunities of current students. The
implementation of DL is a process which uses
available resources and will evolve to incorporate
emerging technologies". (Source: Distance Learning
Distance learning is not a new phenomenon. It
a long history. In 1892, Penn State and
University of Wisconsin were the first universities to
develop a program of correspondence study. The
correspondence model later evolved into the
Independent Learning Program. During the 1960s,
the British Open University in Great Britain
developed off-campus teaching systems, using a
combination of broadcast and correspondence study
systems. Later, many US universities and
universities around the world followed the Open
University model to reach millions of adult
educators. (Source: )
The recent advances in information technology
been responsible for widespread adoption with
distance education. Today distance education has
spread worldwide with extensive use of information
technology with varying degrees of success. Some
programs are very successful in delivering quality
education at a distance and others are not. Quality
distance education needs appropriate mix of
technology, contents, processes, faculty talents while
some demand motivation on the part of the learner.
Now-a-days distance education occurs in a non-
setting where s
tudents participate in
course discussions, exercises, and receive
assessment from the instructors by utilizing
technology such as video conferencing, audio-
graphics, CD-ROM, and Web-based media.
Furthermore, distance learning programs are
becoming increasingly popular at academic
institutions and corporations. Most importantly these
programs are offering learning opportunities for
people that are normally restricted by class time and
M. Islam Y., Rahman M., Rahman Z. and Ashraf M. (2005).
In Proceedings of the Seventh International Conference on Enterprise Information Systems, pages 226-232
space.(Source: Benefits and limitations of Distance
To deliver quality education using available
technology is
the core of distance education
definition. To ensure learning in a distance learning
program several things must happen, e.g. delivery of
contents, regular communication, continuous
feedback, and interactions between the learners
themselves, and also the instructors. Experience
shows that a “live” interaction with the instructor is
the most significant aspect of distance learning
process. Presently there are several ways to achieve
this goal using different available technologies, e.g.,
text chat, audio chat, video conferencing, and
conference calls. Video conferencing and conference
calls have one major limitation. They do not support
“distributed” learning, it is good for only point-to-
point delivery. In some situations the cost of
delivery and the ease of use influence the choice of
technology. In mobile technology, the use of text
messaging (Short Message Service, SMS) is
growing every month. Its rapid growth is generating
substantial commercial applications and research
interest now-a-days. This paper proposes how the
mobiles’ Short Messaging Service or SMS along
with a live telecast can be used to create ‘almost’ an
ideal classroom situation.
The limitation of “distributed” learning is felt to a
much greater degree in countries that are at the
wrong end of the ‘digital divide’. In such countries,
the convenience and facilities afforded by the tools
of information technology are not significantly felt.
In Bangladesh, e.g. empirical estimates of the
number of computers is around one million where
the population stands at over 130 million of which
80% is rural based. Although the number of ISPs
stands at more than 76, most of these are
concentrated in the major cities as are the users. The
estimated number of Internet users is around 13%
(Source: MOSAIC groups report ). Educational
resources in the rural areas are meager with a severe
lack of qualified teachers and educational facilities
Given such tremendous odds, the Government of
angladesh established the Bangladesh Open
University (BOU) in 1992 by an Act of Parliament.
Its prime objective was to transform the country’s
vast human resources into an educated and trained
workforce. The main criterion was to provide access
to educational materials. BOU airs its educational
programs over the national television network. Its
programs are pre-recorded and are run as regular
courses. Students then go to exam centers and get
certificates if they qualify. However, the envisioned
success of an educated and trained workforce is far
from being achieved (official literacy rate: 43%,
Compared with face-to-fa
ce learning, the
limitations of such distance learning are found to be:
¾ Prese
ntations are pre-recorded, i.e. they are not
¾ There is
no interaction between the presenter
and student
¾ There is
no feedback of learning achieved
through a presentation
¾ There
is no monitoring of student progress
throughout the course
¾ Th
ere is no evaluation of teaching quality
To introduce interactivity in the context of
angladesh, we have to look at solutions that would
be technologically acceptable as well as affordable.
In this context, it is seen that the growth of the
number of mobile phones in use has been
phenomenal. As per advertising claims, the total
number of mobile phones in use currently exceeds
3.5 million (Grameen Phone: 2 m, Aktel: 1m, City
Cell + Bangla Link > 0.5 m). The phenomenal
growth of mobile phones is expected to continue
with each mobile company setting their targets in
terms of millions. With new companies like BTTB
and Alcatel joining competition, the unit prices of
calls and sending messages are also expected to
drop. The companies now cover all districts of
Bangladesh. If mobile phone technology could be
used as a tool, it would be much more
technologically suitable for Bangladesh in terms of
the reach it would provide.
The ideal face-to-face classroom situation is when
each student follows the thread and can answer
interactive questions posed by the
lecturer/trainer/teacher. In reality, however, very few
students participate and answer questions posed. If
participating and answering questions, in each class,
could be made a pre-condition for passing the
course, students would have a motive for being
attentive during a lecture. If, for instance, a rule
could be made that a lecturer would ask at least 10
questions during a class session and that each
student would have to attempt at least 8 questions.
To get the credits for the course, each student must
get at least 5 questions right in each class and have
attended at least 80% of the classes. Such a rule
would force attentiveness and participation. This
should also help the learning process. However,
such idealistic rules are meaningless for a face-to-
face classroom situation as it is impossible for a
lecturer to monitor, check and record the
performance of all the students in a large class. In
Distant Learning over national television, it would
be ideal if students could participate and interact
with a “live” lecture (as opposed to a recorded
telecast) just as in a face-to-face classroom. Again,
the vision of such idealistic participation and
interaction is meaningless, as how would the
presenter monitor, check and record the participation
of the large number of distributed students?
3.1 Overview
In the context of Bangladesh the cost of Internet
bandwidth is high and there is no countrywide
internal infrastructure that would enable chat room
technology or video conferencing to be established.
In this backdrop, the mobile telephone industry
appears to offer a feasible alternative. While using
mobiles is still considered costly (e.g. compared
with mobile rates in neighbouring India), the
coverage of the mobile telephones is now
3.2 Methodology
This paper proposes how the mobiles’ Short
Messaging Service or SMS along with a live telecast
can be used to create ‘almost’ an ideal classroom
situation. The participants would of course need
individual mobile sets and access to national
television. The SMS messages would directly
interface with a central SMS server that would
process the messages and show processed output to
the presenter. The SMS server would also respond
directly to the participant as and when necessary.
When required the SMS server would also randomly
dial individual participants so that the presenter can
talk to them directly.
3.3 Operation of a Class
3.3.1 Live Presentation
It is important that the class be conducted “live”, so
that a presenter can ask questions and respond to
clear up wrong concepts. The presenter should also
be able to talk to randomly selected participants. All
the participants should hear this conversation.
Conducting the class on national television would
allow this and a greater outreach. A course or a
country wide training program can be conducted in
this manner.
3.3.2 Class Timing
The class time would be fixed each week and pre-
announced. This would be just like a regular time
tabled class for the entire course or training program.
For example, a course may consist of 10 classes for
one-credit. Each class can be between 50 minutes to
75 minutes. A fixed weekly timing would allow
participants to organize themselves on a weekly
Figure 1: Mobile as an Interactive Tool in Distance
3.3.3 Rules
Along with the class-timing announcement, the rules
of the course would be given. A rule could be that a
lecturer would ask at least 10 questions during a
class and that each student must attempt at least 8
questions. To get the credits for the course, each
student must get at least 5 questions right in each
class in at least 80% of the classes. This would let
the student/participant know what is expected of
him/her. The student can also keep track of her/his
own progress as would the database in the server.
3.3.4 Attendance
When the presenter comes on air for any particular
class, he/she would ask the students to register
attendance for the class. The attending students
would register by sending an SMS to the server. The
server would log the attendance of all the students
who send an SMS. The SMS could contain things
like “y” if already registered for the entire course. If
it is a one-off session, the candidate can register with
full name and other details like date of birth, if
required. The server would maintain the attendance
records. This would be just like a real classroom.
3.3.5 Payment
If a nominal payment is necessary, the payment may
be deducted from mobile “pre-paid” cards. This
would require an agreement with the mobile
company providing the service. This would be a
hassle free payment method.
3.3.6 Lesson
The lesson would be pre-planned and structured in
such a way that questions can be asked at suitable
points every 2 to 5 minutes. The questions would
have multiple-choice answers that would be
displayed as A, B, C and D or A, D, G and J (i.e. the
first letter of the mobile keypad numbers 2, 3, 4 and
5). All the questions for the lesson would be pre-
defined and entered into the computer. This would
force the lesson to be planned with achievable
objectives in mind.
3.3.7 Answering and Monitoring
The number of the question should also be included
such as Q1D, if the chosen answer for Question 1 is
D. The answer would be sent by SMS. The server
would register the answer against the candidates’
phone number or ID and would do three things:
i) The computer would send a message to those who
not attempted the question and tell them their
current status of how many questions have been
attempted. The student/participant would then feel
that he/she is being monitored.
ii) The computer would create a bar chart of the
stions answered and mark the correct answer.
The bar chart would be visible to the presenter and
to the participants. To interact with the participants,
the presenter would select a wrong answer on the
graph and the computer would show those who have
sent the particular answer. The computer would then
randomly select a number that the computer would
subsequently dial. Being connected, the presenter
would ask the student the logic behind selecting the
answer. The presenter on air can then correct wrong
concepts. A few other random selections can be
similarly made. This would help clear up wrong
concepts immediately during the presentation.
iii) A complete record of each student’s performance
hroughout the course would be kept. This record
would be used to issue student’s performance
notification via SMS at the end of a session and also
used to issue certificates at the end of the training or
3.3.8 Question and Answer Session
The last ten minutes of the lesson would be kept for
questions and answers session. The questions will be
typed out as an SMS message sent to the same
server. If anyone asks a question, bonus marks can
be given. The server would log all the questions. The
computer would cluster questions with at least three
or four similar words. The clusters would be visible
to the presenter who would then estimate the nature
of questions and show the estimated question to the
audience. He would then answer the question/s live.
This would allow the students to ask independent
questions and satisfy their needs.
The presenter could also give a task for the next
ass. At the beginning of the subsequent class, a
question can be asked based on the task. Quiz
questions could also be set. Statistical correlations
can be done over a period of time to find whether
students are copying from each other.
3.3.9 Results
At the end of the lesson, each student’s results could
be posted via SMS saying whether the student has
passed or failed the class. The student would have a
record of his/her status. This would motivate the
student to be more attentive in the next class or be
pleased that he/she has done well.
3.4 Costs
The SMS costs per session would be between Tk.20
to 30 at current rates. To promote the sale of mobiles
for this purpose, mobile companies may consider
giving an educational discount of e.g. 50% on the
SMS cost. The mobile companies use single packets
to send each SMS. They can examine the feasibility
of providing an educational SMS service and
become “partners-in-education” against the cost of
3.5 Benefits
The students would get a feeling of participation and
be forced to follow the class. They would feel that
the presenter is monitoring them individually as well
as the coverage would be larger. Countrywide
training of teachers in various schools or health
workers can be conducted in this manner for rural
Bangladesh. The current cost and time of conducting
countrywide training would have to be compared
with the cost of doing it over national television
using SMS technology and a server. The proposed
concept utilizes existing technological infrastructure
to promote learning. There is no need for Internet
and expensive bandwidth. The system would also
lend itself to any type of distributed interactive
session required.
To test out the system, the following software
modules were developed for the SMS Server:
Registration of a course, dates and timing
along with teacher details
Entering questions, multiple choice
answers and correct answer for each class
Registration of a student for a course
Class start reminder system
Attendance at a particular class
Presentation of questions
Logging of answers from each student
Plotting of graphs for each answer
Calculating student responses and
responding to students
Final results module
A lesson was prepared on Quadratic Functions
interspersed with questions. Two practice sessions
were conducted – one drill sequence with the
presenter and one drill session with the sequences of
the SMS server. To send and receive SMSs from the
server a mobile was connected via a data cable to the
server. This was followed by a live session described
Figure 2: Live testing of the Interactive Distance Learning
system with Mobiles
4.1 Live Testing
A class of 52 students was divided roughly 50-50
into those who had mobiles and those who did not.
Both groups sat in two adjacent classrooms, one
face-to-face (f2f) with the presenter and the group
with mobile phones could see the presenter on a
projection screen projected using a video camera and
a multimedia projector as shown in Fig.2.
Both groups were given the same pre-test
and post-test set on quadratic functions. The f2f
group had a pencil and paper to answer the questions
while the mobile phone group used their mobiles to
answer questions via SMS to the SMS server. The
mobile group was given a handout on how to
register and answer questions for the session.
The marks achieved in both the pre-test and
post-test for both groups were put through a t-test.
A separate video camera was used to take
pictures of both the groups in action. It was found
that the group with mobile phones was very excited
about using their mobile phones for this purpose.
Students generally love to use the SMS of mobile
phones. In general it was found that the mobile
phone group on average scored one to two marks
higher than the f2f group. This could be due to the
initial flurry of excitement over the new method
used. It is expected that this excitement may wear
out after the first initial classes.
4.2 T-test
We test the hypothesis that “there is no true
difference between the two means” (NULL
hypothesis). We used statistical t-test for this
purpose.As was mentioned earlier that prior
learning, the pre-test was taken for both groups in
order to validate their merit-similarities. After
lecturing, post-test was taken in order to compare the
results achieved in both learning methods (face-to-
face and SMS-based learning).
4.2.1 Pre-test
In t-test, we got result, h = 0, which means that we
cannot reject the null hypothesis. The significance
was 0.1870, which means that by chance we would
have observed values of t more extreme than the one
in this example in 1870 of 10,000 similar
experiments. The differential means of both learning
was 0.2737. A 95% confidence interval on the mean
was [-0.1387, 0.6861], which includes the
theoretical (and hypothesized) difference of 0.1870.
4.2.2 Post-test
When degree of certainty equals 0.02 or less, the
result, h = 0, means that we cannot reject the null
hypothesis. The significance is 0.0274, which means
that by chance we would have observed values of t
more extreme than the one in this example in only
274 of 10,000 similar experiments. For degree of
certainty=0.02, a 95% confidence interval on the
mean is [-1.9303, 0.0553], which includes the
theoretical (and hypothesized) difference of -0.9375.
But when it is 0.03 or more, the result, h = 1, means
that we can reject the null hypothesis. The results of
the t-test showed that the post-test results of the
Distance Learning class were at least as good as the
f2f classroom.
In (Jackson, R.H. 2001 ) Jackson argued that
technology-enhanced e-Learning is where the
learner audience has the opportunity to meet face-to-
face with the instructor and is a supplement to
traditional, on campus learning. In (Taylor, R. 2002)
Taylor agreed with Jackson and stated that e-
Learning can be used effectively in several different
forms. It can be used as a stand-alone asynchronous
program, or as a synchronous class where all the
students are online at the same time, or as an add-on
to traditional classroom presentations. In
(Alexander, S. 2001) authors further argued that
using technology in both classroom and distance
learning would improve the quality of learning.The
drawbacks of conventional distance-learning is that
it requires more maturity and self-discipline from
students than traditional classroom education, which
may explain the higher dropout rates in distance-
learning programs compared to conventional
programs (Hiltz, S.R. and Wellman, B. , 1997),
(Kumar, A. et al., 2001). A lot of work has been
done in developing prototype-based distance
learning. Some recent Intelligent E-Learning tools
comprising interactive multimedia through Internet
and prototype are illustrated in (Michelle, S. ,
2003)(Peter B., 2004)(Zhang, D. et al., 2004). Our
proposed method overcomes the above problems,
which is validated through questionnaire-based
observation and statistical ‘T-testing’.
The Distance Learning classroom was not in a “real”
rural setting where students would be separated
geographically but would have access to a television
network that would air the educational program.
During the test, the students were sitting side by side
as in a normal classroom. This however brought
home another point. The SMS technology can also
be used in a normal f2f classroom to log and check
answers. The server would track students who are
not answering questions and send them a warning
SMS and remind them of their current status. In the
f2f classroom, students answered questions on a
piece of paper, however, the instructor was not able
to monitor students who were not answering
questions. The mobile as a tool and an input-output
device would be useful in a normal f2f classroom
However, in the last 10 minutes of a class
session, stude
nts can also ask questions via SMS to
the presenter. The computer would group the
questions for the presenter. The cryptic language
used for sending SMS adds another dimension of
difficulty to grouping questions. This part was left
for future work. Research has also to be done on the
kind of courses that can be taught in this manner,
e.g. whether English language can be taught in this
manner. If it can, it would be very useful for places
like Bangladesh where there is a severe shortage of
English teachers.
Distance learning overviews;
Benefits and limitations of Distance Learning; http://
MOSAIC groups report; http:// fac/
Alexander, S. 2001. E-learning developments and
experiences. Ed
ucation + Training. Vol. 43, No. 4/5:
pp. 240-248.
Jackson, R.H. 2001. Defining
e-Learning - Different
Shades of “Online” Web Based Learning Resources
Taylor, R. 2002. Pros and Cons of online learning – a
perspective. Journal of European Industrial
Training. Vol. 26, No. 1, pp. 24 – 37.
Hiltz, S.R. and Wellman, B. Asy
nchronous learning
networks as a virtual classroom. Commun. ACM 40, 9
(1997), 44–49.
Kumar, A., Kumar, P., and Basu, S.C., Student
tions of virtual education: An exploratory study.
In Proceedings of 2001 Information Resources
Management Association International Conference
(Toronto, Ontario , Canada, 2001), 400–403.
Michelle, S. Bachler, Page ; O
ntological Mediation of
Meeting Structure: Argumentation, Annotation, and
Navigation ; 28th July 2003, HTSW 03.
Peter B., “Adaptive e-learning sy
stems: KnowledgeTree:
a distributed architecture for adaptive e-learning”,
May 2004, In the Proceedings of the 13 int
World Wide Web conference on Alternate track
papers & posters.
Zhang, D., Leon, J., Zhao, J., Nunamaker, F.; “Can e-
learning replace classroom learning?”
2004/Vol. 47, No. 5 pp 75-81