LTfLL > overview > 4.2 – The LTfLL Service for Monitoring Conceptual Development (CONSPECT)

4.2 – The LTfLL Service for Monitoring Conceptual Development (CONSPECT)

Download: LTfLL-Service-Description

Problem: In modern educational practice, lifelong learning is a mix of formal and informal opportunities, both of which emphasise development of independent self-directed learning. This is encapsulated by workplace learning environments where learning trajectories reflect interactions of learners with peers and professionals from their own domain, as well as with “clients” (e.g. patients, students, or customers). In such complex circumstances, it is sometimes difficult for learners and their tutors to discern clearly how a specific individual covers key topics and how they might apply this to “real life” issues. Hence, self-directed learning requires support, through formative feedback, but a key issue is how to gather and evaluate the evidence on which such feedback could be based.

 

Solution: CONSPECT is designed to provide a means by which a learner’s conceptual development can be monitored and feedback opportunities are promptly and effectively provided. CONSPECT monitors conceptual coverage of topics based on an automated analysis of textual evidence presented by learners, in comparison with others or over time, to identify shortcomings, misconceptions, and emerging learning opportunities within the learner’s zone of proximal development. It uses textual artifacts from both individuals and groups of learners, such as essays or blogs, to establish a visual model, a “conceptogramme”, of how learners relate concepts to one another. Learners are able to compare their own model with an emerging group reference model in order to identify differences, or to get feedback on where to seek advice from their tutor. This enables learners to monitor their development over time. Tutors can inspect the conceptual development of individuals and groups and use the outputs of the service to inform their interactions with learners.

 

Story: Marion, a learner in the Medical School is on her placement Cardiology for eight weeks. She attends a series of sessions in which she shadows her tutors, observing how they perform their tasks. In the same period, she assesses her competence in diagnosing symptoms and relating these to treatments within predefined PBL cases. This is done in a collaborative setting where she interacts with peers and tutors in online forums. She typically spends time reviewing her previous learning and researching new topics that help her to understand the workplace tasks and the PBL case. As part of her learning process, she reflects on her progressing knowledge and the lessons learnt, maintaining an online journal.

Marion launches CONSPECT, selects the topic space Cardiology and submits her knowledge evidences (from her online reflections and discussion contributions). CONSPECT displays a topic representation based on Marion’s input, showing the identified concepts and their relations. Marion compares her result with three models: that of her peer Peter, an emerging group model, and a tutor defined reference model. Finally she decides to make her model public to feed the emerging group model and to allow others to compare their representations with hers.

 

 

Key functionalities

 

Learner: Enter Topic Space

Go to the topic space URL

View help

View reference model(s)

Create/update personal model

                Add evidence

                View representation

                Save current version

Make current version public (add to data for emergent group model)

Compare earlier version with current

                select earlier version date

                view representation

Compare current version with reference model

                select reference model

                view representation

                get reading suggestions

Compare current version with emergent group model

                select emergent group model

                view representation

 

Teacher: Setup area

Create/update reference model

                Add evidence

                View representation

                Save current version

                 (emphasize most relevant concepts)

Generate topic space

 

Teacher: Enter Topic Space

Select participants & topics:
                select desired tag

                select desired representation

                representations are protected by credential-based access control

Compare topic representations

                Participant vs. reference model

                Emergent model vs. reference

 

Required techniques and data:

Background Corpus

Domain-specific Corpus

Latent-semantic space

Key concepts

Concept clusters (via semantic closeness)

Concept visualiser

Versioning of models

Compare tool for models

Individual model

Reference model

Emerging group model

Input collector/preprocessor

Authorisation

Authentication

 

Specific techniques and data:

Prepare service for a course :

. Data collection

. Data processing

techniques

 

 

Validation

Our validation topics are to investigate the extent to which the CONSPECT:

  • accurately captures what the learners know by the LSA representation
  • identifies to tutors those learners who differ the most from their peers
  • provides the learners more concise, objective, and timely formative feedback

 

We will validate the software in the domain of Medicine in English.  The service uses PubMed to enable interpretation of RSS feeds, for example student journal entries or a discussion board feed. Our validation instruments are interviews with tutors, focus groups with students, questionnaire, and system logging.  Both quantitative and qualitative analysis will be undertaken.

 

tl_files/Ltfll_pics/Gallerie/SUM 4.2.jpg

Link: Get in Touch