slide1 l.
Skip this Video
Download Presentation
Adaptive Educational Environments for Cognitive Skills Acquisition

Loading in 2 Seconds...

play fullscreen
1 / 183

Adaptive Educational Environments for Cognitive Skills Acquisition - PowerPoint PPT Presentation

  • Uploaded on

Adaptive Educational Environments for Cognitive Skills Acquisition. Ashok Patel Director, CAL Research De Montfort University Leicester, United Kingdom Tel/Fax: +44 116 257 7193 Kinshuk Information Systems Dept, Massey University Palmerston North, New Zealand

I am the owner, or an agent authorized to act on behalf of the owner, of the copyrighted work described.
Download Presentation

PowerPoint Slideshow about 'Adaptive Educational Environments for Cognitive Skills Acquisition' - gin

Download Now An Image/Link below is provided (as is) to download presentation

Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.

- - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - -
Presentation Transcript

Adaptive Educational Environments for Cognitive Skills Acquisition

Ashok Patel

Director, CAL Research

De Montfort University

Leicester, United Kingdom

Tel/Fax: +44 116 257 7193


Information Systems Dept, Massey University

Palmerston North, New Zealand

Tel: +64 6 350 5799 Ext 2090 Fax: +64 6 350 5725



  • Adaptive educational environments for cognitive skills in applied domains
  • Accommodation of both the ‘instuction’ and ‘construction’ of knowledge
  • Tutorial aims to provide better understanding of design principles for development of such environments
  • Design based on informed educational methodologies


Who are we?


Major use of adaptive educational environments

Life long learning

Life long learners


Life-long “learning”

  • Continuous non-formal and informal learning activities (during and beyond student life!)
  • Motivation to learn is typically problem focused
  • Observation and practice are key elements.
  • Starts from birth!!

Life-long “learners”

  • Includes practitioners who have to learn continuously due to rapid changes in the technology employed for their work (just-in-time, learning-on-demand)
  • Learners generally do not have direct contact with subject experts
  • Learners traditionally pursued distance learning avenues.

Life-long learning and computers

  • Absence of human experts requires some kind of ‘filling the gap’
  • A range of contemporary technological hardware, software and systems are sought to impart the domain competence to the learners
  • What is domain competence!

Domain competence


Domain Knowledge


Skills (Cognitive + Physical)


Constituents of Domain Competence

Reflection oriented and abstract

Action oriented and experiential

Easier to

learn from mistakes

Difficult to

learn from mistakes



Trial and error

logical processes




An example of the know-how aspect of know-when is the temporal context required for an appropriate sequence of operation

An example of the know-why aspect of know-when is the environmental and behavioural contexts required for making a decision


Context oriented and both experiential and abstract



Awareness oriented


Constituents of Domain Competence


Ä It has an operational orientation.

Ä It is mainly action-driven and hence pre-dominantly experiential.

Ä It is difficult to inherit it from someone else’s experience.


Ä Learning from mistakes.

Examples : Computer simulation and virtual reality


Constituents of Domain Competence


Ä It has a causal orientation.

Ä It is mainly reflection-driven and therefore based on abstraction.

Ä It can be inherited from someone else’s line of reasoning.


Ä Logical processes.

Ä Needs deeper reflection.


Constituents of Domain Competence

Know-when (and -where)

Ä It has a contextual orientation.

Ä It provides the temporal and spatial context for both the know-how and know-why. It is thus both action and/or reflection driven.


Constituents of Domain Competence


Ä It has an awareness orientation.

Ä It includes above three types of knowledge in terms of know-what.

Ä It also contains information about the environmental context of this knowledge.


How about computers?


Which constituent of knowledge can be successfully facilitated through computers?

Know-how Know-why

Know-when Know-about


Constituents of Domain Competence

Know-how (skill based) is easier to learn from making mistakes. It requires less reflection and hence most suitable for computer based learning.


Instruction in knowledge context

Ideally, an instructional system, designed for novice users, teach all knowledge constituents.

But, know-why is difficult to handle mainly for two reasons:

1. It needs natural language interaction.

2. It needs use of metaphors, which are difficult to understand for a novice user.

Know-how, on the other hand, is operational, and can be conveyed to the user more easily, even with symbolic representations.


Instruction in knowledge context

Traditional hypermedia based ITSs approach, in general, has been to teach the know-why aspect of knowledge with the help of explanations.

The links provide stimulus to the user to know more about a particular topic.

System works more as a friendly librarian and learning depends on the initiative of a student.


TLTP Byzantium approach

  • TLTP Byzantium has used symbolic representations to explain the concepts by taking advantages of precise and concise mathematical notations.
  • Various features of the approach are:
    • Bottom-up approach
    • No centralised knowledge
    • Each Intelligent Tutoring Tool (ITT) is free standing

The Byzantium principles

  • use computers and human for what they are (currently) good at
  • employ useful software tools in the overall learning environment consisting of human teachers and education technologies

The Byzantium principles

  • add applied intelligence to the software tools to provide a degree of support to students, enabling them to work by themselves
  • understand the economics of the learning environment and be concerned with assessment and course management as they consume substantial human resource

Knowledge levels

1. Introductory application level, where the formation of a mental map of interrelated conceptual atoms takes place and the students learn how to use the basic tools of a discipline.

2. Advanced application level, where the vertical and horizontal integration takes place.

3. Actual application approximation level, where an attempt is made to stimulate the real world situation and students learn to understand the behavioural and environmental contexts of the procedural knowledge.


Intelligent Tutoring Tools (ITTs)

  • Mixed-initiative systems
  • All the knowledge is stored in two-fold knowledge base
  • Not intended to replace the human tutor, but to supplement
  • To be mixed and matched with other technologies (e.g. video)
  • Suitable for various configurations such as classroom, open and distance learning

Structure of an ITT

A network of inter-related variables where the whole network remains constant.

Example, partial network of 7 out of a total of 14 variables in marginal costing.


Student-ITT Interaction

  • No fixed sequence of filling in the variables.
  • Any value can be entered in any variable, provided the whole network remains consistent.
  • Correct values are accepted even if the intermediate values are missing.
  • On entry of a wrong value and missing intermediate values, the student is advised to carry out necessary intermediate step(s).

Granularity Negotiation

  • Provision of intermediate variables in the interface.
  • Multiple screen interface where one screen carries forward the aggregation of the details in the previous screen.
  • Functional interface that zoom-in to show the constituents of a complex value presented as a single variable on the main interface e.g. formula in the Capital Investment Appraisal ITT.

Structure of an ITT

Knowledge Base

1. Variables

2. Relationships

3. Tolerances


- Student

- Lecturer

- Administrator

Inference Engine

Expert Model

1. Correct values

2. Derivation procedure

(Local expert model)






Student Model




1. Student input

2. Value status (filled or blank)

3. Derivation procedure

4. Interface preferences


Input (student answer, position)

(four levels)



User Interface


Context based link to textual description


Lecturer’s model answer to any lecturer generated narrative questions

(Remote Expert Model)


1. Calculator

2. Table Interface

3. Formula Interface


Application specific


Tutoring Strategy of an ITT

  • Introduction of complexity in phased manner
  • Corrective, elaborative and evaluative aspects of student model are used for tutoring.
  • Learning process is broken down to very small steps through suitable interfaces.
  • ‘Road to London’ paradigm is adopted to eliminate the need for diagnostic, predictive and strategic aspects.

Feedback from tutoring module

Context sensitive messages are generated to improve semantics and to prevent monotony. There are several possibilities:

  • If the student input is correct
  • If the student input is incorrect:
    • If the value can be derived directly from the work done by the student and/or from the given information
    • If enough information isn’t available to directly derive the value


  • Independent evaluation of Byzantium and another CAL package at University of Glasgow over three years period (Stoner & Harvey, 1999)
  • Student performance improved significantly since introduction of CAL
  • Improvement appeared to be mainly reflected in the students’ ability to complete numeric questions.

Evaluation: Student Feedback

“Byzantium was useful because you could go over bits you were unsure about. It was better than a book because it was interactive. With the interactive questions you tend to pay more attention than you would to a book.”

“Byzantium offers instant feedback, is more involving and you can do as many questions as you like.”


Evaluation: Student Feedback

  • 71% students showed a preference for Byzantium material while 8% indicated no particular preference.
  • The students wanted more tutoring systems like Byzantium, observing that it was good to use CAL if the tutoring software was good.

Future: Integrating the ITTs

The ITTs can be integrated in two directions:

  • Vertical integration will allow holding and comparing results of different instances of an ITT, e.g. comparing four different springs.
  • Horizontal integration will allow use of multiple ITTs to solve a given problem, e.g. design of a typical safety valve using spring design, screw design and other ITTs.

Integrating the ITTs

An individual ITT is, thus, an autonomous entity, connected in a decentralised network to make more sophisticated tutoring system.

The integration can be achieved either:

using an intelligent advanced application level interface


by software agents who would approach the ITTs searching for a solution of the problem in hand.


Domain Skills

Senso-motoric skills are externally visible. They can be easily acquired by visualising the processes and learn through observations.

Cognitive skills run inside human mind. Their non-visible nature demands a more sophisticated learning process.

Life long learners have more difficulty in acquiring cognitive skills!


What is the best suitable theoretical framework for facilitation of cognitive skills?


Cognitive apprenticeship framework

  • Modelling: Learners study the task pattern of experts to develop own cognitive model
  • Coaching: Learners solve tasks by consulting a tutorial component of the environment
  • Fading: Tutorial activity is gradually reduced in line with learners’ improving performance and problem solving competence

Phases of Cognitive apprenticeship

World knowledge (initial requirement)

Observation of interactions among masters and peers

Assisting in completion of tasks done by master

Trying out on own by imitating


Phases of Cognitive apprenticeship

Getting feedback from master

Getting advise for new things on the basis of results of imitation, comparing given solution with alternatives

Reflection by student, resulting from master’s advice


Phases of Cognitive apprenticeship

  • Repetition of process from 2 to 7
    • Fading out guidance and feedback
    • Active participation, exploration and innovation come in
  • Assessment of generalisation of the tasks and concepts learnt during repetition process

Cognitive apprenticeship


Guided discovery

(guidance/instruction; observation of demonstrations to understand the concepts, exploration within limited boundaries, guidance/feedback …)


Example systems

  • Cognitive apprenticeship based learning environment (CABLE)
  • Interactive simulations based learning environment (InterSim)

CABLE objectives

Environment should facilitate:

  • acquisition of basic domain knowledge;
  • application of the basic domain knowledge in non-contextual and contextual scenarios to get skills of the discipline; and
  • generalisation of the domain knowledge to get competence of applying it in real world situations.

CABLE architecture

  • Observation - for acquisition of the concepts
  • Simple imitation - skills acquisition through articulation of the concepts
  • Advanced imitation - for generalisation and abstraction of already acquired concepts and for acquisition of skills of applying concepts in different contexts

CABLE architecture

  • Contextual observation - for deeper learning after imitation process results into the identification of gaps in learner’s current understanding of the domain knowledge
  • Interpretation of real life problems - for acquiring competence in such narrative problems as encountered in real life situations

CABLE architecture

  • Mastery in skills - for repetitive training
  • Assessment - for measurement of overall progress


Observation phase



Interactive learning and testing phase

Icon showing Interactive learning mode

Icon showing

system generated problem mode: generated values given in grey

Problem Space:

Students have to fill in the remaining values

Control Panel

Interactive messages:

dialogue, feedback and guidance



Fine grain interface and contextual help



Context based problem creation facility



Future work on mental process modelling


Analytical disciplines focus on domain knowledge while task oriented disciplines emphasise more on skills.

Task oriented medical profession requires core knowledge of medical concepts, senso-motoric skills in physical tasks (such as surgery), and cognitive skills in diagnosis of diseases and other decision making tasks.


InterSim learning environment is an exploratory learning environment with individual intelligent assistance geared towards facilitating core knowledge and cognitive skills.


Educational framework

  • InterSim environment adopts Cognitive Apprenticeship framework (Collins, Brown & Newman, 1989) to support domain knowledge while focusing on skills acquisition. Accordingly, the environment supports following activities:
  • Modelling
  • Coaching
  • Fading

Characteristics of InterSim environment

  • The InterSim environment supports:
  • learning in and outside educational institutions where the constraints of space, time or resources limit the possibility of regular institutional access
  • different configurations of learning such as classroom based, open and distance learning.
  • acquisition of cognitive competence in both knowledge and task oriented performance.

InterSim system for Ear domain

Aims: Competence acquisition by the learner in basic and advanced knowledge of the domain and associated cognitive skills.

Target audience: Medical students and doctors in continuing medical education.

Additional features: Knowledge sharing among doctors by adding and documenting real cases of the domain within the system.


Main Functional States in InterSim

Learning AssessmentCase Authoring

Immediate (dynamic) feedback on user actions with a view to facilitate active learning.

Coarse grained instruction dominated learning:Basic knowledge of subject matter and overview of the domain with the help of tutoring and guided tours.

Fine grained knowledge construction:Advanced understanding of specific areas with the help of intelligent interactive simulations.


Main Functional States in InterSim

Learning AssessmentCase Authoring

Cognitive skills development:In a constructivist manner through repetitive learning.

Application of the acquired knowledge and skills:Identifying and correcting any misconceptions acquired or gaps left in earlier learning, with the help of domain problems, including notoriously difficult cases recorded in the field, for the learner to attempt diagnosis and solution.


Main Functional States in InterSim

Learning Assessment Case Authoring

  • Assesses the knowledge and skills acquired within the Learning State without providing any assistance.
  • Assessment of acquisition and retention of the knowledge and skills
  • Assessment by comparing learner’s solution with master solution.
  • Delayed (static) feedback

Main Functional States in InterSim

LearningAssessment Case Authoring

Authoring facilities for doctors/teachers to add real life cases.

These cases can be accessed for learning and assessment purposes.

Teachers can employ the documented cases to provide case-based teaching away from the system, if so desired.


Intelligent assistance

Intelligent assistance and adopts adaptive behaviour according to learners’ needs.

The system adopts exploration-space-control (ESC) methodology to facilitate the exploration of learning space with reduced cognitive overload.

Within the Assessment State, ESC assists in selecting suitable problems to match the individual learner’s level of learning.


Dynamic messages

Dynamic messages adopted to the situation and learners’ current needs.

Handling messages:

Context adaptive navigational aids to help learners identify various interactions available in the system. Increases amount of ‘useful exploration’.

Learning messages:

Subject domain related messages, driven by the learner interactions, based on learner’s state of knowledge and competence.


What exactly we mean by



Adaptive Educational Systems?



Increased user efficiency, effectiveness and satisfaction


Improved correspondence between learner, goal and system characteristics


Need for “Intelligence”/adaptivity

  • Users generally work on their own without external support.
  • System is used by variety of users from all over the world.
  • Customised system behaviour reduces meta-learning overhead for the user and allows focus on completion of actual task.

Adaptable Systems

Systems that allow the user to change certain system parameters and adapt the system behaviour accordingly.

Adaptive Systems

Systems that adapt to the users automatically based on system’s assumptions about user needs.


How does adaptivity work?

  • System monitors user’s action patterns with various components of system’s interface.
  • Some systems support the user in the learning phase by introducing them to system operation.
  • Some systems draw user’s attention to unfamiliar tools.
  • User errors are primary candidate for automatic adaptation.

Levels of adaptation

  • Simple: “hard-wired”
  • Self-regulating: monitors the effects of adaptation and changes behaviour accordingly
  • Self-mediating: Monitors the effects of adaptation on model before putting into practice
  • Self-modifying: Capable of chaging representations by reasoning about the interactions

Problems in adaptation

  • User is observed by the system, actions are recorded, giving rise to data and privacy protection issues.
  • Social monitoring becomes possibility.
  • User feels being controlled by the system.
  • User is exposed to adaptation concept favoured by the designer of the system.
  • User may be distracted from the task by sudden automatic modifications.

Recommendation for adaptive systems

  • Means for user to (de)activate adaptation
  • Offering adaptation in the form of proposal
  • User may define specific parameters used in adaptation
  • Giving user information about effects of adaptation hence preventing surprises
  • Editable user model

Content based adaptation

Multiple Representation Approach


Multiple Representation Approach

  • Adaptation for multimedia based systems
  • Presents multimedia objects (such as audio, pictures, animations) into a multimedia interface world where the relationships of the objects to the world are governed by the educational framework.

Components of MRA

  • Multimedia object selection
  • Navigational object selection
  • Integration of multimedia objects

Components of Multiple Representation Approach

Multimedia object selection

  • Task specificity and learner’s competence
    • Different multimedia objects are suitable for different tasks
      • Audio is good to stimulate imagination
      • Video clips for action information
      • Text to convey details
      • Diagrams are good to convey ideas

Components of Multiple Representation Approach

Multimedia object selection

  • Task specificity and learner’s competence
    • Level of learner’s domain competence in the current situation should be considered
    • Curriculum should follow a granular structure to allow assessment on individual units
    • This will ensure context based selection of multimedia objects

Components of Multiple Representation Approach

Multimedia object selection

  • Task specificity and learner’s competence
    • Granularity in domain content in two dimensions:
      • Advancement in curriculum (e.g. initially an abstract concept using animation of a concrete instantiation of the concept, followed by more complex abstract representation)
      • Details of the content (.g. static diagram at novice level, and VR with full complexity at advanced)

Components of Multiple Representation Approach

Multimedia object selection

  • Task specificity and learner’s competence

Components of Multiple Representation Approach

Multimedia object selection

  • Expectations
    • Expectations of learner and of domain about representation of the tasks should be considered
    • If they don’t match, possibly provide presentations in more than one form to suit all expectations (e.g. learner wants overview by graphic, but domain requires textual details, give both)

Components of Multiple Representation Approach

Multimedia object selection

  • Reference & revisits of already learned domain content

“revisiting the same material, at different times, in re-arranged contexts, for different purposes, and from different conceptual perspectives is essential for attaining the goals of advanced knowledge acquisition”.

(Spiro et at., 1991)


Components of Multiple Representation Approach

Multimedia object selection

  • Reference & revisits of already learned domain content
    • MRA favours revisiting the same domain content in different contexts.
    • Use of similar multimedia objects is favoured since it puts less cognitive overload on the user.

Components of Multiple Representation Approach

Multimedia object selection

  • Reference & revisits of already learned domain content
    • enforces links between current concept and the referred one
    • enhances the mental model of previous concept and its generalisation in multiple situated scenarios
    • ease in learning current concept by making familiarisation with past learning experiences

Components of Multiple Representation Approach

Multimedia object selection

  • Use of multi-sensory channels
    • Adequate use the visual, aural and tactile senses of the learner.
    • Chances of getting distraction due to the unused channel are high.
    • Reception enhances if the representation of domain content involves various sensory channels.

Components of Multiple Representation Approach

Multimedia object selection

  • Context based selection of multimedia objects
    • If multiple multimedia objects are available for same task/concept, presentation should use the most suitable object in the context (e.g. interaction possibility by simulation to beginner, whereas review of concept in textual form by experienced learner
    • Demand of domain should determine which multimedia object is required for which task and in which context.

Components of Multiple Representation Approach

Multimedia object selection

  • Authenticity of multimedia objects
    • The learner should be aware of the authenticity of the multimedia objects (Laurel et al., 1992) (e.g. appropriate messages to the learner while showing schematic diagrams and animations of the processes that do not show the real objects)

Components of Multiple Representation Approach

Navigational object selection

  • Navigation in typical educational systems is via links
  • A link does not say what happens to the screen when the user activates the link (Rada, 1995)
  • Learner’s expectations of outcome while activating a link should be matched with the presentation of actual resulting interface connected to the link (either static as in traditional systems or adaptive/dynamic as in intelligent systems).

Components of Multiple Representation Approach

Navigational object selection

  • Type of link should suit to the context and learner’s expectations towards the outcomes.
  • Links should be used for the tasks for which they suit best and do not put cognitive overload on the learner.
  • Selection of links should not deviate learner’s attention from the main task of learning.
  • The existence of link should be as transparent as possible.

Components of Multiple Representation Approach

Navigational object selection

  • MR approach favours the use of both, interaction objects (e. g. push buttons, radio buttons, check boxes) and interactive objects (e.g. text, pictures) (Bodart & Vanderdonckt, 1994)
  • Interaction objects provide transition from one part of the system to another on learner’s explicit initiative
  • Interactive objects facilitate a contextual transfer recommended by the system

Components of Multiple Representation Approach

Navigational object selection

  • Types of navigational links:
  • Direct successor

leading to the successive domain unit in knowledge hierarchy.

Such transfer should arise from current context such as link in text or message after fulfilling learning criteria of current domain unit.


Components of Multiple Representation Approach

Navigational object selection

  • Types of navigational links:
  • Parallel concept link

leading to the analogous domain unit for comparative learning


to the unit related to another aspect of currently being learnt domain content.

These transfers should be explicit (using interaction object).


Components of Multiple Representation Approach

Navigational object selection

  • Types of navigational links:
  • Fine grained unit link

leading to the fine details of the domain content once some missing or mis-conceptions are identified

These transfers are very contextual and it is necessary to maintain the context during transfer.

Interactive objects such as image maps should be used.


Components of Multiple Representation Approach

Navigational object selection

  • Types of navigational links:
  • Glossary link

leading to a pop-up “spring loaded” module (Nielsen, 1996) in exploration process

Available only till learner is interested in it and explicit action required (such as pressing the mouse button).

These links provide a referential summary of the terms, hence should be initiated from the terms themselves.


Components of Multiple Representation Approach

Navigational object selection

  • Types of navigational links:
  • Excursion link

leading to a learning unit outside the current knowledge hierarchy, within current context

A way back to main learning unit should be possible

The context for excursion links should be broad enough to cover the essence of current unit


Components of Multiple Representation Approach

Navigational object selection

  • Types of navigational links:
  • Problem link

leading to the problems related to current conceptual unit.

Transfers to problems should result from system’s inference of learning criteria fulfilment of a conceptual unit.


Components of Multiple Representation Approach

Integration of multimedia object

  • No more than one active multimedia object at a time on the screen (except comparative studies)
  • Integration of multimedia objects should be complimentary and synchronised
  • Same material should not be repeated using different multimedia objects at the same time

Components of Multiple Representation Approach

Integration of multimedia object

  • Integration of decision intensive objects is not recommended due to their high cognitive load demand
  • To avoid confusion, different multimedia objects not initially distinguishable should not be put together
  • Integration of dynamic and static observation objects should be such that both objects should not use same sensory channel at the same time

Components of Multiple Representation Approach

Integration of multimedia object


User exploration based adaptation

Exploration Space Control



  • Exploration is a self-initiated learning activity.
  • Learning by exploration is an effective technique for task oriented disciplines such as computer science and medicine.
  • It provides not only skills of the domain but also the understanding of the embedded concepts.

Learning by exploration

  • Hypermedia systems, simulation systems and other similar systems provide learners with exploration environments where they can explore various paths to solve problems.
  • Learning actually takes place by accessing various information resources such as hypertext, demonstrations, simulations, and so on.

Learning by exploration

  • Exploration activity: searching these information resources to comprehend the information and to acquire domain concepts/knowledge.
  • Comprehension and acquisition involve mutually integrating the information from different resources, and integrating new information into existing knowledge.

Exploration space

Extent of the exploration activity


Extent of the information resources (including the domain concepts/knowledge)


Exploration operations (such as search, selection, apply, integration etc.)

This is called exploration space.


Problems in educational exploration

  • Learners should be free to explore to “construct” their learning (Carroll et al.,1985).
  • However, learners may not know what to and how to explore.
  • Excessive mental efforts to search and integrate the information from different information resources may cause cognitive overload.
  • Exploration space may also be quite wide so they may lose their ways.

Need for adaptation in exploration

The extent and amount of complexity inherited in existing exploration techniques should depend on the learner’s current level of competence and their current capacity to cope with cognitive load in such explorations.


Examples in exploration adaptation

  • Adaptation of navigation the exploration paths that learners should follow
  • Tailoring the information to be presented to the learners

(making it easier for the learners to search and comprehend domain concepts and knowledge)


Examples in exploration adaptation

  • Various parameters are restricted in simulation-based learning systems to make it easier to interpret the results.
  • Some systems sequence the problems in such ways which focus learners’ attention on specific parts of the domain. This allows an easier understanding of the domain in gradual manner.

Supporting learning by exploration

  • The exploration space needs to be limited for the novice learner, and restrictions should gradually be removed as the learner progresses in the learning process.
  • Exploration Space Control (ESC) is the over arching phenomenon encompassing these adaptive mechanisms.

Exploration space control (ESC)

  • ESC controls the extent of exploration space according to domain complexity and to the learners’ competence, understanding levels, experiences, characteristics, etc.
  • It is also a technology that integrates current technologies for exploration.

Exploration space control (ESC)

  • ESC facilitates proper learning environment for all types of learners.
  • Several tools are provided through suitable user interface to explore the exploration space.
  • Exploration tools and information is restricted according to the learners’ attributes.
  • However, these restrictions would be as less as possible, and would be reduced with learners’ progress of subject matter understanding.

Purposes of ESC

To facilitate active learning:

Suitable for learners with higher learning competence. Achieved by reducing cognitive load as less as possible.

To facilitate step-by-step learning:

Suitable for learners with lower learning competence. Achieved by reducing cognitive load as much as possible.

A combination of ‘active learning’ and ‘step-by-step learning’ covers whole learner spectrum, and therefore ESC is applicable for all kinds of learners.


ESC Control Levels

  • Embedding information: This facilitates the creation of information space and involves scaffolding.
  • Limiting information resources:
    • Limiting number of information resources
    • Selecting types of information resources appropriate for looking into current domain material

ESC Control Levels

  • Limiting exploration paths:
    • Limiting the number of feasible exploration paths to be looked into
    • Limiting the exploration paths which are non-feasible or are unrelated to the current domain material
  • Limiting information to be presented:
    • Limiting the amount of information.
    • Adapting the contents of information to each learner

ESC and Current Technologies

Control Levels

Embedding information

Limiting information resources & exploration paths

Limiting exploration paths

Limiting presented information

Limiting exploration paths & presented information

Current Technologies



Problem ordering (courseware)

Information tailoring

Simulation setting


Designing systems with ESC

  • Identification of learning goals to be accomplished by the learners
  • 2. Selection of scaffolding methods
    • Selecting various information resources to accomplish each learning goal (e.g. hypertext, simulation, demonstration)

Designing systems with ESC

  • 2. Selection of scaffolding methods
    • Developing the information resources (decision about the information to be presented) based on:
      • Amount of the information
      • Contents of the information (such as abstract/concrete, detail, and theory/example)

Designing systems with ESC

  • 2. Selection of scaffolding methods
    • Selecting various exploration operations to be used in and between each information resource (such as Select, Trace, Apply, Integrate, and Interpret).

Designing systems with ESC

3. Deciding levels of control to be applied to different information resources:

a) Deciding the major purpose of ESC (active or step-by-step learning)

Helps designers to decide on the ways of how to control exploration space


Designing systems with ESC

b) Deciding control levels


Designing systems with ESC

  • Deciding application of control levels according to learner and domain models
  • Learner model factors:
  • Preferences
  • Knowledge Levels
  • Experiences
  • Competence
  • Exploration Process
  • Cognitive Load (Mental Efforts)
  • Domain model factors:
  • Type of knowledge (know-how, know-why …)
  • Degree of detail (Granularity)
  • Depth (Deep or Shallow)

Pedagogical issues in the development of adaptive educational systems


leading to

Human Teacher Model

(not just for cognitive skills based systems)


Research in Context

Intelligent systems:

  • Knowledge management
  • Reasoning
  • Natural language processing

Intelligent tutoring systems:

  • Knowledge representation
  • Discourse management
  • Other architectural aspects to improve student computer interaction and to provide effective tutoring strategies

Current implementation of context is only within the systems.

There are various contexts outside the system that have an

important bearing on the success of an ITS!


Contexts of intelligent tutoring systems

Besides the inter-actional context, the environmental and objectival contexts are important for any educational system.


Inter-actional Contexts

  • Employed within the system to provide an improved human-computer interaction.
  • They facilitate more intelligent feedback by the system.
    • Plan recognition
    • Knowledge structuring
    • Knowledge representation
    • Reasoning
    • Discourse management

Environmental Contexts

  • The major constituents of environmental contexts:
    • Student (capabilities, preferences and motivation)
    • Teacher (preferences and outlook)
    • Discipline (nature)
    • Knowledge (characteristics)
    • Medium (capabilities of computer hardware and software)
    • Social and technological environment

Environmental contexts


An innovative use of technology may require uncomfortable scrutiny of educational purpose and teaching method.

- Hammond & Trapp


Environmental contexts

ITS in a joint cognitive system

Traditional teacher-student interaction is a complex phenomenon, affected by personality, background, motivation and host of other factors and the same is also true for the peer-to-peer interaction.

An ITS that intervenes in this rich environment needs to demonstrate intelligent behaviour, not only in its interactions with a student but also in its interactions with a teacher.


Environmental contexts

The role of a teacher

“... don’t try to mimic the way people communicate, just try to design the system so it complements human communicative skills”

- Devlin

The role of the teacher stands out as a partner within the 'joint-educational system'. An ITS should understand this role and help a teacher rather than prematurely attempt to act as a replacement.


Environmental contexts

Teaching styles and students

High workloads, surface level assessment demands and lack of freedom in learning environment force students to use reproductive approaches. Does a teacher’s teaching style also get affected similarly?

If a teacher is forced to adopt a superficial teaching style due to external factors, the situation can be improved by harnessing ITS in a supportive role to free up some of the pressure.


Environmental contexts

Support for a teacher within an ITS

While it is worthwhile within the design laboratories to integrate advanced pedagogical strategies of modelling, coaching, reflection, articulation, scaffolding and fading as well as exploration within an ITS, it is also practically worthwhile for an ITS to support a teacher in adopting these strategies in a limited way but with a degree of ‘intelligence.’


Environmental contexts

Human teacher model

We suggest that a human teacher model should formally be incorporated in the design of an ITS to recognise the different teaching styles, to record the teaching styles adopted in the design and preferably to enable adaptation to suit the implementing teacher.


Environmental contexts

Teaching styles and the discipline

An explicit explanation of the teaching style will not only enable an implementing teacher to understand the designer’s rationale but will also help in dealing with the cognitive dissonance arising from any differences in the teaching styles.


Environmental contexts

Nature of discipline

A discipline’s subject matter and the degree to which its practice is regulated contributes to the way it is taught.

Computers currently are good for the disciplines where learning is based on ‘doing’ and ‘observing’ rather than ‘freely conversing’ and ‘arguing’.


Environmental contexts

Levels of learning a discipline & representations

At the introductory conceptual level, learning benefits more from graphics, animations, audio and video clips as they provide multiple stimulus to improve motivation and facilitate concept acquisition.


Environmental contexts

Levels of learning a discipline & representations

Procedural learning benefits from direct manipulation of symbols, text and graphical representations for some disciplines.

While others benefit from analysis, exchange of opinions and expert feedback preferably in a synchronous system such as video conferencing and electronic white boards or asynchronously using email or discussion forums.


Environmental contexts

Levels of learning a discipline & representations

Theoretical learning benefits from some form of hierarchical representation enabling smooth granular transitions.


Environmental contexts

Characteristics of the medium

Technologies such as hypertext, multimedia, hypermedia and virtual reality offer increasing ease and flexibility in knowledge construction by the learners, however, they are not an unmixed blessing. They are accompanied by an additional cognitive load and a potential for distractions.


Environmental contexts

Characteristics of the medium

The novice learners should benefit from richer representations as they provide multiple stimulus.

The same learners may most likely get distracted in the absence of directed learning as they may not have developed adequate meta-cognitive skills of setting learning goals, selecting effective learning techniques, monitoring progress towards goals, and adjusting strategies as needed.


Environmental contexts

Characteristics of the medium

  • Instructional methods and not the media cause learning.
  • Human brain, product of millions years evolution, is not changing rapidly and can be overloaded by the sensory output that technology is capable of delivering.
  • - Clark
  • To prevent such overloading & distraction, amount and richness of contextual information may have to be constrained.

Environmental contexts

Characteristics of the medium

A representation that is efficient from the learning point of view may be inefficient for a performance task and the optimisation of efficiency and expressiveness is often mutually exclusive requiring a trade-off, possibly per domain.

- Dutton & Conroy

System design should be governed by needs of discipline and not by some abstract model or emerging technologies.


Environmental contexts

Social & cultural effects

“Designing interfaces for culturally diverse users is fundamentally a problem of communicating the intended meaning of representations ... in every culturally determined usability problem a divergence between the target meaning and the interpreted meaning of representations was present.”

- Bourges-Waldegg & Scrivner


Environmental contexts

Social & cultural effects

Their study found, however, that direct intercultural communications between users are less problematic since the users develop jointly a communication space in order to succeed in their task, despite differences in culture and language.


Environmental contexts

Social environment

“So far in the research literature, little attention has been given to the effect of class or socio-economic differences upon variations in learning style.”

- Anderson

Differences in backgrounds, goals, or outlooks on life can be problematic in communication between two people every bit as much as their not speaking the same language.


Objectival contexts

Teaching and Assessment

  • Concerned with “teaching-assessment objectives anomaly”
  • Human teacher attempts to cover whole syllabus, the assessment procedure does not require to perform 100%.
  • Is the objective of educational system to encourage acquisition of the facts and rules or is it to encourage acquisition of meta-learning abilities?

Contexts of an ITS

The success of an ITS depends on adequate consideration of the various contexts encompassing its design and implementation. While there is an increasing recognition of context in the 'Intelligent' aspect of an ITS, there is a need for recognition that context affects the 'Tutoring' and 'System' aspects as well.


Modelling human teacher

  • Human teachers may have:
  • different personalities
  • different teaching styles (born out of their traditional, progressive or vocational outlook and their own learning style)

It is not possible to envisage all the preferences of implementer teacher at design time.


Human Teacher Model

We recommend a re-configurable human teacher model to be incorporated in the design of ITS:

  • to recognise the different teaching styles
  • to put on record the teaching style(s) adopted in the design, and
  • enable manual or automatic adaptation to suit the implementing teacher

Why explicit record of designer’s teaching style(s)

  • better understanding of designer’s rationale by implementing teacher
  • help in dealing with the cognitive dissonance arising from any differences in teaching styles

Why explicit record of designer’s teaching style(s)

  • clear rationale behind adopted teaching strategy may also help in the student learning in less adaptive systems
  • easier understanding of representations which are difficult due to cultural differences
  • if designer’s teaching style is unproductive in a culture, the system may be localised

Summary: Incremental growth

  • ITS, in their current stage, cannot replace all the functions of human teacher.
  • Efforts should be on increasing productivity (just like initial word processors for steno-typists).
  • ITS designers should treat human teacher as their target user.
  • Human Teacher Model is next logical approach in that direction.

Designing cognitive skills based adaptive educational systems

Multiple representation approach

Exploration space control


MRA in InterSim system

Multimedia object selection

  • Learning process starts with receptive and active observation using:
    • Text and Normal static pictures for receptive observation
    • Sensitive pictures for active learning in current domain hierarchy
    • Semi-sensitive pictures for active learning outside of current domain hierarchy.

MRA in InterSim system

Multimedia object selection

  • Advanced observation takes place using three types of animations:
    • Automatic
    • user controlled
    • user initiated.

MRA in InterSim system

Multimedia object selection

  • Following observation phase, simulations are used for acquiring skills.
  • More realistic learning environment is provided by pictorial virtual reality.
  • Even more realistic cases are provided by video.
  • Decision making skills are provided by flowcharts.

MRA in InterSim system

Navigation object selection


MRA in InterSim system

Integration of multimedia objects

  • Examples:
  • The concept of "appropriate sound energy routing " is presented by two comparative animations.
  • Structure of ossicles is presented as alternative static picture and pictorial VR to prevent confusion due to similar initial visual states.

MRA and student modelling

  • To act as filter on student model

Content recommended by student model to present to the learner

MRA Validator

validation not ok

Content manipulator


MRA Renderer

Interface presentation to student


MRA and student modelling

  • Functionality achieved:
  • XML Schema for Multiple Representation Approach implementation.
  • Stylesheets based on XML Schema to transform XML notation into web pages.
  • Formalisation of limited multimedia objects and link types.
  • Products: 1. Manual validator
  • 2. Limited renderer

MRA and student modelling

  • Next steps
  • Customisation interfaces for rules modifications and objects addition/deletion
  • Automatic verification compiler stylesheets for dynamic domain content presentation based on student model state
  • Content manipulator to resolve validation problems
  • Fully functional renderer
  • Prototype demo

ESC in InterSim

  • Determination of learning goals:
  • Goal a) Understanding the structure and functionality of human ear
  • Goal b) Acquiring appropriate skills in diagnosing and treating ear diseases.

ESC in InterSim

2. Scaffolding

a) Selection and development of various information resources:

Hypertext:description of structure and functionality of ear (goal a)

Demonstration:behaviour of normal ear (goal a)

Simulation:experiment on functionality of ear (goal a), exploration for diagnosis and treatment of diseases (goal b)

Problem ordering:problem sequencing based on learner competence (goal a), enabling learners to solve (diagnose and treat) problems in order (goal b)


ESC in InterSim

2. Scaffolding

b) Deciding control levels to restrict exploration operations

Hypertext:Select & Trace (for goal a)

Demonstration:Interpret (for goal a)

Simulation:Apply & Interpret (for goals a & b)

Problem ordering:Trace (for goals a & b)

Design:Select, Apply & Interpret (for goal b)


ESC in InterSim

3. Deciding control levels

a) Major purpose of ESC in InterSim system (to provide active learning)

b) Selection of control levels

- Limiting information resources (for goals a & b)

- Limiting exploration paths (for goals a & b)

- Limiting presented information (for goal a)

- Embedding information (for goal b)

c) Deciding how to control various information resources


Example of Control Levels in InterSim

Concept of main path and excursions


Main learning path: Structure of middle ear

Excursion: Physics of sound

Limitations: Only sound travel through mechanical linkage of ossicles is presented. Information regarding sound transfer from air to water (related to inner ear) and other similar information is not presented to maintain the context.


Example of Control Levels in InterSim

Limiting information resources for understanding the structure and functionality of ear

- Restriction methods: Restricting the representation of domain material in terms of complexity of representation (for example, static pictures vs virtual reality scenarios)

- Deciding parameters: Exploration experience, Cognitive load


Example of Control Levels in InterSim

Limiting exploration paths for understanding the structure and functionality of ear

- Restriction methods: Restricting buttons, combo box choices, anchors/ links to be used in explorating hypermedia to limit Select & Trace operations.

- Deciding parameters: Exploration competence, Knowledge level


Example of Control Levels in InterSim

Embedding information for acquiring skills diagnosing and treating diseases

- Restriction methods: Providing scaffolding so as to decrease domain complexity with regard to learner models (for example, first allowing the learner to semi-explore the disease development process in an animation wizard; then adding simulation capabilities to allow the full exploration; then adding extra simulation capabilities for diagnosis)

- Deciding parameters: Exploration competence


Future of Exploration Space Control

The Implementation of Exploration Space Control has provided various adaptive features towards the learner and thus validated the applicability of the methodology for systems providing learning-by-exploration.

Further research demands implementation of ESC in other system providing learning by exploration to assess the generalisation of the methodology.


Exploration Space Control Research

  • Next steps
  • Development of framework to enhance student model based on “Exploration Space Control” methodology
  • ESC Validator
  • ESC Content Manipulator
  • ESC Renderer


Cognitive skills based adaptive educational systems may prove very effective in rapidly increasing life-long and just-in-time learning process if:

  • they are based on appropriate educational framework
  • contexts surrounding their design process are considered appropriately
  • adaptation is provided both at content and user level