Help:Concept Maps

This page describes the concept maps in detail.

Overview of the Concept Maps
Concepts are an important innovative feature of the reference model.

Simply put, a concept is a generic e-assessment word or phrase; for example are all concepts. Moreover, concepts are semantically interconnected; one concept can have several other concepts - called "subconcepts" - related to it. There are two types of concept supported by the reference model
 * “peer mark”
 * “interoperability”
 * “question purchaser”
 * “accessibility”


 * process concepts – which relate to e-assessment functionality
 * entity concepts – which relate to e-assessment data

All concepts, together with their interrelationships, are encapsulated in and rendered by two graphical concept maps. These maps allow reference model users to


 * orientate themselves within the overall e-assessment landscape, by comparing their personal knowledge and practice with the concepts identified


 * navigate the domain at different levels of abstraction, by “drilling down” through concepts’ interrelationships


 * retrieve actual resources stored in the model. Concepts are generic implementation-neutral terms, which are linked to the resources stored in the reference model. For example, the “Deliver” node on the process map is associated with all resources (e.g. the BS7988 standard) which logically relate to the delivery of e-assessments.

Development of the Concept Maps
It is crucial that the reference model is useable – to anyone interested in e-assessment. A user should be able to interact with the model in an intuitive way, and gain immediate value from doing so.

How will a typical user initially connect with, and subsequently navigate through, the vast volume of data held inside the reference model? This holds especially true for newcomers to e-assessment, which is a vast and constantly evolving subject area.

The approach adopted was to provide a high-level implementation-neutral language interface which reflects the common parlance of e-assessment. This approach allows the user to latch onto a familiar word or phrase, and to start his or her journey through the reference model from there.

How did the project team develop this interface? First we reviewed existing models of the e-assessment domain. However, none was sufficiently high level or covered the area as extensively as we sought.

Thereafter the team adopted a standard software engineering technique; it performed both a functional and a data-oriented analysis of the underlying domain (the team comprised e-assessment practitioners of many years’ experience). This crystallised a set of generic e-assessment terms, or “concepts” in FREMA parlance. The term “concept” was coined from the use of a concept-mapping tool to record the results of the analyses.

Two concept maps were developed in tandem (see below). The following points relate to the development of these maps:-
 * an iterative process of concept development took place over approximately six months. Several project meetings reviewed the latest version of the concepts maps, resulting in further extensions and refinements
 * each map applies to both manual and computer-based systems. Indeed, a particular environment may implement some functions (e.g. Delivery) wholly via software, yet implement other functions (e.g. Process Appeals) entirely manually
 * there is no implied sequence of nodes in either map; the first order nodes are arranged alphabetically starting at the 12 o’clock position
 * there was much debate over some apparently simple terms e.g. “feedback” and “specification”. This reflects the overloading of many terms in e-assessment
 * it was deemed useful to incorporate external definitions wherever possible; for example, the set of roles is taken from the “Ultimate Assessment Engine” paper [UAE]. This approach adds provenance to the FREMA concepts
 * there was always an inclination for “just one more" refinement of the maps; however the project timescale imposed a final end date for this activity
 * during their development, the concept maps were exposed to the UK e-assessment community for comment e.g. at the JISC – CETIS Conference “The e-Framework: Priorities and Challenges for 2006” Heriot-Watt University, Edinburgh, 15 /16 November 2005

The Process Concept Map
This map reflects the generic processes which occur within e-assessment

The Entity Concept Map
This map reflects the generic data components which are used within e-assessment

Interactive Implementations of the Concept Maps
The FREMA project team developed interactive animated representations of the concept maps.

Each has been implemented using MacroMedia Flash.


 * Entity Concept Map
 * Process Concept Map

How to Interpret the Concept Maps
Each concept map provides a particular structure perspective on the e-assessment domain - indeed they are two different representations of the same space.

Each concept map has one central node, from which radiate the first level concepts. The latter represent the highest level (i.e. coarsest grained) concepts which the model supports.

For example, the process map contains the following fourteen first level concepts



Refining Concepts via Subconcepts
In turn, a first level concept may have one or more semantically related other concepts – called “subconcepts”.

A subconcept refines its parent concept in more detail (i.e. it is finer grained than its parent). A subconcept is physically connected to its parent concept via a straight line. Note that a second level concept can itself have one or more subconcepts at the third level – and so on.

Note that subconcepting is an optional mechanism; however it is recommended to identify as many subconcepts as possible.

We will look at two subconcept examples.

1. The “Manage Users” first order process concept has two subconcepts – “individual” and “group”.




 * The parent concept “Manage Users” represents a high-level e-assessment process. It represents all tasks involved in administering users within a typical e-assessment environment.  ( The definition of this concept is: Set up individuals and groups in the system, edit their details and delete them.' For medium or high stakes assessment, it is necessary to maintain secure login names and passwords for participants. These details must be kept up-to-date and there must be a secure mechanism for changing passwords. One requirement may be to allow 'anonymous' access to sit an assessment, which according to [UAE], 'will not receive feedback from markers'.


 * The “Individual” concept explicitly relates to a subset of its parent’s tasks – i.e. those involved in looking after individual users. (The definition of this concept is "Register candidates, authors and other users with the system. Edit and delete users as necessary").


 * The “Group” concept explicitly relates to a subset of its parent’s tasks – i.e. those involved in looking after groups of user. (The definition of this concept is "Assign individiduals to groups so that tests can be scheduled for particular groups").

2. The “Interaction Type (IMS QTI)” first order entity concept has four subconcepts.

Each of these refines the generic concept (i.e. an interaction type) into a particular type of interaction (e.g. “text based”).



In summary, subconcepts allow the user to navigate through and retrieve data from the model at different levels of abstraction.

How to Use the Concept Maps
The concept maps are accessible from the “Home” menu on the top left hand corner of any FREMA screen.

Each map has an “Instructions” link at the top left hand corner of its surrounding frame. Click this link to display the following information



You can perform the following screen operations on a concept map

Move the entire map on screen
Each map is rendered within a fixed frame; therefore the more nodes you display, the more “crowded” the screen will appear. Click and drag the entire map to focus on a particular node and its children

Single click a concept node
This event will cause two side-effects on screen


 * 1) A yellow border will appear around the selected concept node
 * 2) Extra information will appear on the right hand portion of the screen.



This is concept indexing at work - all entities associated with the selected concept node appear as a list on the right hand side of the screen. Use the vertical scroll bar to navigate this list.

Note this list shows entities related to both the selected node and all its subconcepts.

Warning: The interface allows you to select more than one concept node at a time. Deselect a selected node by single clicking on it again – the yellow border will disappear. Make sure you only have exactly one concept selected before you examine the list of entities on the right hand side.

Double click a concept node
You may double click any concept node which has a blue triangle next to it.


 * 1) If the triangle points downwards, double clicking the concept will display all its subconcepts.
 * 2) If the triangle points upwards, double clicking the concept will hide all its subconcepts.

Note that, if screen space is scarce, the rendering software will temporarily 	grey out 	overlapping background nodes.

Rest mouse cursor on a concept node
Whenever the cursor rests on a concept node, a definition of that concept will 	appear in a 	pop-up frame, as illustrated below.



Alter the size of the map
Use the mouse wheel, or the slider bar control at the top left hand corner of the map’s frame, to enlarge or reduce the size of the map.

How to Map Concepts against Entities
Concepts act as an indexing mechanism into the resources referenced inside the reference model. If you are authoring your own information into the model, you will need to know about mapping concepts to resources.

There is a many-to-many relationship between resources and concepts. In other words
 * each resource is mapped to (potentially) many concepts
 * each concept is mapped to (potentially) many resources

How, as a reference model contributor, do you implement these mappings? The answer is via associations. An association is a link between two resources in the model. A concept is itself an resource, and therefore can participate in associations.

Creating associations between concepts and resources

 * First consider the resource you wish to “conceptualise” (e.g. the BS7988 standard)


 * Next examine both concept maps, and identify as many lowest level concepts within each you think most accurately describe the resource at hand. Take a note of this list of concepts. In this example, you could map BS7988 to the Specification node on the Entity map – that would suffice.  However, it is better to map it instead to the National subconcept of Specification - thereby providing more specific information about it.

Go to the associations screen. Click the “Create new entry in ‘associations’” button. Enter the following information
 * Remember: The following instructions apply to just one of the concepts you identified above; you will therefore have to repeat these steps if you have identified two or more relevant concepts.

Source Object: An entity in the database (e.g. the BS7988 standard)

Destination Object: A concept (e.g. “national”)

Note: following diagram is from "old" interface - will have to be updated.




 * Add a description for the association


 * Click the “Add entry” button

Inheritance of associations
Creating a single association as above may generate other associations “behind the scenes”. This will occur if the chosen entity has subconcepts.

Consider the following concept tree.



C1 has two subconcepts – C2 and C3. C2 has two subconcepts – C4 and C5. C5 has one subconcept – C7.

Consider creating an association between the concept C5 and the entity E. The above mechanism would create the following three associations within the reference model.

•	E  C5 (i.e. that created on screen as above) •	E  C2 (automatically created) •	E  C1 (automatically created)

An association is automatically created for each parent concept of C5 – working backwards towards a first order concept (in this case C1).



Thereafter, if a user selects C1 or C2 or C5 on the appropriate concept map, the entity E will appear on the list of associated entities.