DESIGNING ELEARNING MODELS

The Agora Framework

Jos´e Rangel-Garc´ıa

and Jorge Buenabad-Ch´avez

Edusistemas, D.F., Mexico City, Mexico

Departamento de Computaci´on, CINVESTAV-IPN, D.F., Mexico City, Mexico

Keywords:

eLearning, Modelling, Framework, Software Engineering, eLearning Engineering.

Abstract:

This paper introduces the Agora Framework (AF), a modelling tool to help design, evaluate and deploy eLearn-

ing models. An eLearning model in AF resembles a business model, and consists of various blocks, each block

addressing one main aspect of the eLearning business, such as: learners (consumers), eLearning types (prod-

ucts) to be delivered, pedagogical basis for course design, hardware and software resources, among others.

We describe how to use AF in designing an eLearning model, and outline the role that AF may play within an

eLearning engineering.

1 INTRODUCTION

The beneﬁts of eLearning go without saying. Most

potential users and deployers of eLearning are willing

to adopt it for academic or training purposes. eLearn-

ing is now being deployed in various ways, from sim-

ply sharing ﬁles in a web server to serious games on-

line, or from a single course to academic programmes

with ofﬁcial value. The simplest types of eLearning

are obviouly easy to deploy and practically risk-free,

but complex types may become a rather difﬁcult and

costly project, and even unsuccessful.

Currently, most deployers of relatively complex

types of eLearning are large institutions who can

afford to take the risk. This is in contrast to the

wide adoption, proportionally, of web systems across

small, medium and large enterprises and institutions.

That is, small and medium institutions, e.g., private

schools, have not yet adopted eLearning systems as

widely as they have adopted web systems, despite the

fact that both kinds of systems can be deployed with

free open software and have many aspects in com-

mon: web (HTML) pages, multimedia content, hier-

archical menus, and on-line content delivery, among

others. We realise that not all of eLearning takes place

through the Internet and Web technologies, but most

of it does nowadays; and we assume this from now

on.

The main reason for the less widely adoption of

eLearning systems is that they are a relatively new

class of web systems. As such, they have unique as-

pects that need to be addressed with new, tailor-made

best practices, methods and tools, i.e., an eLearning

engineering. The basic principles of an eLearning en-

gineering will be those of software engineering:

• Clearly deﬁned goals and requirements

• Systematic development of the (eLearning) sys-

tem in phases

• Careful planning of these phases

• Continuous audit of the entire development pro-

cess

There is a lot of published work that is related

to eLearning engineering, but most of this work is

addressed to a particular aspect of eLearning. The

term eLearning engineering is still rarely used, and

where used (eLearning Engineering, 2012; Lischka

and Karagiannis, 2004; Toval et al., 2011), no com-

prehensive view of the subject is discussed. A com-

prehensive view of eLearning engineering is clearly

essential for the wide adoption of eLearning; in order

for small and medium, and even large, deployers of

eLearning to better assess, decide and cope with the

issues they will face in adopting eLearning.

New disciplines do not emerge overnight, even in

the fast-paced ﬁeld of ICT. eLearning engineering, as

a scientiﬁc discipline towards engineering-type devel-

opment of eLearning systems, will take time to ma-

ture. It is likely to develop in a similar way Web En-

gineering developed. Web engineering (Pressman and

Lowe, 2008; G. Kappel and Retschitzegger, 2006)

265

Rangel-Garcia J. and Buenabad-Chávez J..

DESIGNING ELEARNING MODELS - The Agora Framework.

DOI: 10.5220/0003956902650270

In Proceedings of the 4th International Conference on Computer Supported Education (CSEDU-2012), pages 265-270

ISBN: 978-989-8565-06-8

 2012 SCITEPRESS (Science and Technology Publications, Lda.)

was developed for the purpose of efﬁciently address-

ing the particular issues of web systems. It borrows

heavily from the long-standing knowledge of soft-

ware engineering (Bell, 2005). As software engineer-

ing, web engineering advocates best practices, meth-

ods and tools to develop a feasibility study, require-

ments engineering, planning, ..., and project manage-

ment, but tuned for web systems development. For in-

stance, central to web systems is the display of infor-

mation (textual and/or multimeda), and thus web en-

gineering includes methods and tools to organise, dis-

play and update the information. In particular, Press-

man and Lowe (Pressman and Lowe, 2008) advocate

the use of agile methods for web systems develop-

ment; characteristics of agile methods include: incre-

mental delivery of software functionality, customer

involvementthroughout the development process, and

embracing change.

From our experience in developing eLearning sys-

tems, we suggest that the main stumbling block to

adopt them is the great difﬁculty to specify “clearly

deﬁned goals and requirements”. One factor for this

is the many concepts and issues involved. eLearn-

ing systems are a combination of education, comput-

ing, text and multimedia content, technology, soft-

ware development, and platforms. The development

of eLearning systems should therefore be perceived

as a multidisciplinary approach requiring knowledge

and expertise from different areas. Another factor is

the lack of a method and tools to specify clearly de-

ﬁned goals and requirements.

This paper proposes the Agora Framework (AF)

to address the two factors just mentioned. AF is a tool

and a method to help generate e-learning models. An

eLearning model does not correspond to a high-level

design of an eLearning system. Rather, it is similar

to a business model (Osterwalder and Pigneur, 2010),

comprising the main elements involved in delivering

a particular type of eLearning. An eLearning model

includes a cost structure, goals, learners (consumers),

eLearning types (products) and resources, among oth-

ers. In deﬁning an eLearning model, stake holders

implicitly arrive at “clearly deﬁned goals and require-

ments”.

AF is based on the ideas about business model

generation in (Osterwalder and Pigneur, 2010). The

rationale behind its design is that content in eLearning

systems (i.e., course content) is a product on and by

itself, as opposed to only information about a product

or service as is the case in other web systems.

Section 2 describes the elements AF. Section 3

describes how AF should be used. Section 4 outlines

its role within a tentative eLearning engineering. We

conclude in Section 5.

2 THE AGORA FRAMEWORK

The Agora Framework (AF) is a diagram composed

of nine blocks/elements/modules, which represent the

main components of an eLearning model (business)

whose purpose is to deliver (sell) eLearning. The nine

blocks are shown in Figure1.

As a diagram, the AF has the purpose of helping

to: i) visualise all the elements involved in running

an eLearning model; ii) classify possible embodiment

options for each element, e.g., a particular type of

eLearning to be delivered; and iii) reason about each

element and its relationship to other elements, i.e.,

how different options of one element affect the op-

tions of other elements, e.g., how different resource

conﬁgurations affect cost. We mention below the

main issues for each element of AF, about which in-

formation must be gathered and analised in order to

make informed decisions in designing an elearning

model.

Learners Segments

Students are the core (the clients) of an eLearning

model. Satisfying their learning needs is the pur-

pose of any eLearning system. To better analyse such

needs, it is convenient to classify the students into

segments according to some criteria, such as: com-

mon needs, per (external) company/institution, basic

skills, etc. An eLearning model can deﬁne as many

segments as needed, but must deﬁne them precisely in

order to facilitate the analysis of their needs, current

level of relevant knowledge, skills, attitudes, preju-

dices, and skills related to ICT.

As a guiding principle, a group of students should

conform a segment if their learning needs require a

particular conﬁguration of any of the blocks in the AF,

thus giving rise to a particular eLearning model.

Once learners segments are deﬁned, and their

needs well understood, the eLearning model can be

designed considering those needs.

e-Learning Types

We deﬁne an eLearning type as a particular organi-

sation/conﬁguration of all Agora elements selected to

fulﬁll the learning objectives of a particular learners

segment. An eLearning type must be designed and

developed. It may be innovative, using new methods

and technologies; or it may be typical, using known

technologies but with new or more features and at-

tributes. The guiding principle during design is that

design makes the difference. eLearning may be as

CSEDU2012-4thInternationalConferenceonComputerSupportedEducation

266

Figure 1: The Agora Framework.

good as the best in-the-classroom (face-to-face) learn-

ing, or as bad as the worst in-the-clasroom learning.

A good design of an eLearning type starts with

a good instructional design (Dick, 1996; Clark and

Mayer, 2008; Horton, 2006). The instructional de-

sign advocates selecting, organising and specifying

the learning experiences needed to teach something

to someone. In addition, the instructional design:

• drives and supports decision making in the selec-

tion of content and everything related to learning

material and technologies to use, e.g., learning

management system (LMS).

• drives the development of content and media se-

lection.

• determines budget decisions.

Media

The media in the classroom used to be only symbols

in the blackboard and the voice of the teacher. Today

we must design the media: the adequate mixture of

text, graphs, voice, music, sound effects, animation

and video. Then, we must design a sequence of them

and synchronise it with the delivery medium, such

as Internet, intranet/extranet (LAN/WAN), audio- and

video-tape, satellite, interactive TV, CD-ROM, etc.

What media is best? Research comparing differ-

ent media suggests that it is not the media what brings

learning about, but the instructional methods (Clark

and Mayer, 2008). That is, if the instructional meth-

ods are essentially the same, the media does not mat-

ter. However, each media offers unique capabilities to

deliver content, and new innovative ways of deliver-

ing content should be explored to capitalise on those

capabilities (Wigdor and Wixon, 2011).

Designers should also consider technologies with

which students are familiar and feel comfortable, such

as mobile phones, PDAs, social software, email, etc.

Goals

The Goals block contains the goals of the eLearning

model, i.e.: what is to be accomplished by the organi-

sation that is developing the eLearning model. Need-

less to say how important it is to clearly deﬁne such

goals. A simple way to proceed is as follows:

1. Establish the goal of, what is most important

for, the sponsoring organisation. Is it proﬁt, public

service, reputation, or ROI?

2. Describe how the project will contribute di-

rectly to achieve the goal of the organisation. If you

cannot convincinly and honestly describe/argue how

the project will contribute to achieve the organisation

goal, abandon the project; the organisation will not

be interested and will not support you. Otherwise you

will have resources and support. Pass to next step.

3. Deﬁne learning objectives. Describe how stu-

dents will be changed through the eLearning sys-

tem to be developed. Learning objectives may have

prerequisites, which in turn may become objectives

in themselves having prerequisites, and so on. You

should identify the basic skills of potential students.

Examples of organisation goals for eLearning

models are the following:

• Reducing cost of education in 50% in one year.

• Making a proﬁt of $ 200,000.00 out of selling

courses.

• Training and certifying 150 plant operators.

• Increasing material covered in a course in 10%.

• Deploying a platform to administer courses and

improve information exchange between teachers

and students.

Pedagogical Basis

This block contains the pedagogical principles that

should drive the design of learning content, and this

has an effect on all the other blocks in the framework.

DESIGNINGELEARNINGMODELS-TheAgoraFramework

267

A good pedagogical design ensures that no incon-

sistencies exist between the curriculum being taught,

the teaching methods used, and the evaluation pro-

cedures adopted. One should ﬁrst clearly deﬁne the

learning results to achieve; then select the learning

activities that will offer students a good opportunity

to achieve the learning results; and ﬁnally, design the

evaluation assignments that will genuinely show that

learning results have been achieved.

Key Resources

Every eLearning model requires resources to achieve

the eLearning results expected. Resources may be

physical, content, human, owned by the organisation,

rented, or provided by third parties.

Physical resources: classrooms, technological in-

frastructure, computers, networks, learning manage-

ment systems, application software, software tools,

etc.

Content resources: text, images, audio, videos,

etc. These resources are not easy to develop, and are

costly. They are the product to sell.

Human resources. In eLearning models, human

and content resources are closely related, and both

are key factors in one of the most difﬁcult problems

to solve: content production. Teachers may lack the

skills or the tools, or may not know how to use tech-

nology with pedagogical principles, to generate effec-

tive content.

We describe below how two courses of differ-

ent complexity are typically developed and the most

likely outcome with regard to quality and resource us-

age.

a) Course of low complexity. A teacher creates

and formats herself content: text, ﬁgures, question-

naires and assignments. She also assembles the con-

tent, e.g., within an LMS.

b) Course of high complexity, requiring more peo-

ple:

• Teachers contribute knowledge and basic mate-

rial.

• Tecno-pedagogues design teaching material ac-

cording to pedagogic and didactic principles and

around available technological tools.

• Producers use various applications to create the

teaching material: text, images, audio, video, an-

imations, questionnaires, assignments, etc. Then

assemble the content within an LMS.

Typically, in case (a) the content produced is of

low quality, and the eLearning resources are under-

used. Also, the content is usually boring to students.

The teacher may become overloaded during develop-

ment. In case (b), eLearning resources are better used,

but the cost and time spent is a lot more; many organ-

isations cannot afford the cost.

Using AF, these issues will be brought to light,

and in case (a) it will help determine the amount of

training required by teachers, and in case (b) to ﬁnd

alternatives to reduce cost.

Key Activities

Key activities are designed to make learning happen-

ing. People learn through pondering, investigating,

analysing, evaluating, organising, synthesising, dis-

cussing, testing, deciding and applying ideas. Learn-

ing activities should promote these kind of mental

processes. Learning activities can be classiﬁed thus:

Absorption. The purpose of this type of activity is

to convey knowledge to the student, typically reading

text, watching a video or listening a narrative. The

student is mentally active but not physically.

Action. In this type of activity the student learns

while doing something: playing, answering ques-

tions, or following a procedure. The student practises,

explores and discovers.

Connection. This type of activity drives stu-

dents to relate what they are learning to their previ-

ous knowledge; it facilitates applying knowledge later

when needed.

Key Providers

It is not always possible for an organisation to acquire

all the resources or carry out all the activities involved

in an eLearning model. Working with third parties,

e.g., outsourcing, costs can be substantially reduced.

eLearning service providers offer a range services

including: serious games, consultancy, ad hoc course

development, hosting of LMS, real-time eLearning

solutions, among many others.

Cost Structure

The Cost Structure block will describe the most im-

portant costs related to the development and manage-

ment of the eLearning model. Each block of AF in-

volves a cost, which may be of one of the following

kinds:

Fixed Costs: constant costs independently of the

level of operation of the eLearning system, i.e., num-

ber of students using the system.

Variable Costs:.variable costs on account of the

level of operation of the eLearning system.

Economy of Scale: cost advantages derived as the

CSEDU2012-4thInternationalConferenceonComputerSupportedEducation

268

Figure 2: AF before use and in use.

output/use of the eLearning system grows. E.g. host-

ing larger eLearning systems cost proportionally less

than smaller systems.

The cost of an eLearning model should not be

more nor less than is required to accomplish the learn-

ing objective, i.e.: that the students do what they

are expected to do. In practice, the optimum cost

is achieved gradually through successive approxima-

tions.

However, during the design of the eLearning

model, relevant cost information should be collected

and analysed so as to have a good estimate from start.

3 USING AGORA FRAMEWORK

As mentioned earlier, AF has the purpose of helping

to: i) visualise all the elements involved in running

an eLearning model; ii) classify options within each

element; and iii) reason about the relationships be-

tween elements. Figure 2 shows AF before use and

in use. Legends in the ﬁgure correspond to: Learner

Segments (LS), e-Learning Types (ET), Media (MD),

Goals (GL), Pedagogical Basis (PD), Key Resources

(KR), Key Activities (KA), Key Providers (KP), and

Cost Structure (CS). AF should be drawn on a black-

board or printed large and pinned to a wall.

However, before using AF, the following activities

should be made:

• Gather and analise information about each ele-

ment of AF (LS, ET, ... CS), particularly current

solutions, methods and tools, etc.

• Organise various meetings inviting relevant peo-

ple, e.g., teachers, experts in education, cloud

computing solutions, eLearning platforms, etc.

The use of AF consists in discussing, analysing

and classifying the information gathered onto each el-

ement of AF

, as shown in Figure 2. These discussing

and classifying in itself generates various eLearning

models. The best one, under some criteria and con-

straints, will eventually be chosen.

Note that meetings should carry on until the par-

ticular aspects of each element of AF are well deﬁned

and understood by the people most closely related.

For instance, it should be very clear how teachers will

participate in course content generation, and if teach-

ers themselves need training.

It is also worth noting that, if each element of

AF is sufﬁciently deﬁned, then the chosen eLearning

model would comprise most of the information typi-

cally found in a Requirements Engineering document.

This information would only need to be organised and

extended as required. A Requirements Engineering

document tipically consist of: 1. Functional Require-

ments, 2. Data Requirements 3. Performance Re-

quirements 4. Constraints 5. Guidelines (Bell, 2005).

A feasibility study usually precedes the require-

ments engineering process in order to establish

whether or not the project is worth proceeding. This is

typically established through a cost-beneﬁt analysis:

“the cost of providing the system vs the money saved

or created by using the system – the beneﬁt” (Bell,

2005). Using AF, a feasibility study is implicitly car-

ried out throughout the deﬁnition of the eLearning

model, as this includes a cost structure (CS).

4 PUTTING IT ALL TOGETHER

In this section we outline an entire process to design

and deploy an eLearning model, and where AF ﬁts

within this process. We believe something similar to

this process will be part of an eLearning engineering.

As a reference, we ﬁrst list the typical phases found

(suggested) in a software engineering process (Bell,

2005); web engineering process are similar (Pressman

and Lowe, 2008):

Feasibility Study; Requirements Engineering;

Planning; User Interface Design; Architectural De-

sign; Detailed Design; Programming; System Integra-

tion; Validation and Veriﬁcation (testing); Production

(deployment); Documentation; Maintenance. Project

management runs along the entire process after the

Planning phase.

Agora is a greek word that means forum; by the term

Agora Framework we mean: a framework to be used (to

design eLearning models) in a forum-like way.

DESIGNINGELEARNINGMODELS-TheAgoraFramework

269

An eLearning Engineering: eLearning Model

Design and Development

The eLearning model design and development pro-

cess we propose has ﬁve phases: Prepare, Collect, De-

sign, Develop, and Manage. The progression through

these phases is rarely lineal. In particular, the Collect

and Design phases tend to proceed in parallel. The

last phase, Manage, is about continuously managing

your deployed eLearning model.

1. Prepare: Describe the motivation behind the

project. Establish a common language to describe, de-

sign and analyze and discuss eLearning models. Se-

lect the design team. Prepare the place for discussion.

2. Collect: Research the elements for the eLearn-

ing model design: learners, technology and environ-

ment. Collect information, interview experts, and

identify needs and problems.

3. Design: Transform the information and ideas

from the previous phase into eLearning models using

the Agora Framework. After an intensive eLearning

model inquiry, select the most satisfactory eLearn-

ing model. Create the eLearning model requirements

document/s.

4. Develope: Plan for developing and deploying

the project; Develop the required software; Get the

needed hardware and software; Hire the required ser-

vices, Train teachers and administrators; Produce the

content for learning; Perform the prototype phase (if

it was speciﬁed); Deploy the complete eLearning sys-

tem.

5. Manage: The deployed eLearning model

should be placed in a continuous feedback loop so

that managers can identify and change the parts of the

process that need improvement. A method such as the

Deming cycle could be used:

PLAN: Design or revise eLearning process com-

ponents to improve results.

DO: Implement the plan and measure its perfor-

mance.

CHECK: Assess the measurements and report the

results to decision makers.

ACT: Decide on changes needed to improve the

process.

5 CONCLUSIONS

eLearning offers many beneﬁts to users and deploy-

ers, but developing eLearning systems is currently

difﬁcult and costly (except for simple types of eLearn-

ing) because many new issues are involved and it is

little understood how to proceed. An eLearning en-

gineering is needed to guide the adoption of eLearn-

ing without too much risk. The purpose of the Agora

Framework (AF) proposed here is to help achieve

“clearly deﬁned goals and requirements” of eLearn-

ing models. AF helps to think of and visualise the

issues related to an eLearning project, and their re-

lationship. We believe AF will help non-technical

people to understand what is involved in an eLearn-

ing project. We have outlined its use and how it ﬁts

within a whole process of adopting eLearning. We

are currently working on designing a methodology to

generate requirements documents out of an eLearning

Agora model.

REFERENCES

Bell, D. (2005). Software Engineering for Students. Pren-

tice Hall, fourth edition.

Clark, R. and Mayer, R. (2008). e-Learning and the Science

of Instruction: Proven Guidelines for Consumers and

Designers of Multimedia Learning. Pfeiffer, San Fran-

cisco, 2nd edition.

Dick, W. (1996). The dick and carey model: Will it sur-

vive the decade? In ETR&D, Vol. 44, No. 3, ISSN

1042-1629, pages 55–63. Association for Educational

Communications and Technology (AECT).

eLearning Engineering (2012). Engineering e-learning. In

www.elearning-engineering.com. Visited on 29 Jan,

2012.

G. Kappel, B. Proll, S. R. and Retschitzegger, W. (2006).

Web Engineering: The Discipline of Systematic De-

velopment of Web Applications. John Wiley & Sons,

San Francisco, new edition.

Horton, W. (2006). e-Learning by Design. Pfeiffer, San

Francisco, new edition.

Lischka, J. and Karagiannis, D. (2004). A generic e-

learning engineering framework embracing the se-

mantic web. In Artiﬁcial Intelligence: Methodology,

Systems, and Applications, 11th International Confer-

ence, AIMSA 2004, Varna, Bulgaria, September 2-4,

2004, Proceedings, pages 341–350. Springer.

Osterwalder, A. and Pigneur, Y. (2010). Business Model

Generation. John Wiley & Sons, New Jersey, new

edition.

Pressman, R. and Lowe, D. (2008). Web Engineering: A

Practitioner’s Approach. McGraw-Hill Higher Edu-

cation, USA, new edition.

Toval, A., Toval, R., de Gea, J. M. C., Nicolass, J., and

Fernandez-Aleman, J. L. (2011). Learning systems

development using reusable standard-based require-

ments catalogs. In 2011 IEEE Global Engineering

Education Conference (EDUCON): Learning Envi-

ronments and Ecosystems in Engineering Education,

pages 907–912. IEEE.

Wigdor, D. and Wixon, D. (2011). Brave NUI World: De-

signing Natural User Interfaces for Touch and Ges-

ture. Morgan Kaufmann.

CSEDU2012-4thInternationalConferenceonComputerSupportedEducation

270