2013 03 14 (educon2013) emadrid ucm elearning standards learning analytics

E-Learning Standards E-Learning Standards and Learning Analyticsand Learning Analytics

Can Data Collection Be Improved by Using Standard Data Models?

IEEE EDUCON Conference 2013 Berlin, March 14th, 2013

Ángel del Blanco [email protected]

2/15

OutlineOutline

Learning Analytics

E-Learning Standards (GBL)

Learning analytics + e-learning standards

3/15

Learning Analytics: a tendency on the rise!Learning Analytics: a tendency on the rise!

4/15

Students can use these analysis results as guidance and self-awareness tools;

Teachers can use them to identify issues and try to tackle them;

Schools can use results as a domain-specific variant of Business Intelligence

to detect and address learning problems, assess students, and

predict learning results

5/15

Learning Analytics: Source of dataLearning Analytics: Source of data

Collect, report, predict, act and refine

Learning Management Systems (LMS): Wide adoption

Lots of different tools

MOOCs Increasing acceptance

Store large amounts of data about the students’ performance..

6/15

Learning Analytics: Issues gathering dataLearning Analytics: Issues gathering data

LMSs lack standardized data structures;

LA tools tend to be tied to specific implementations of LMS and databases.

This has a number of negative consequences: data gathered across different LMSs, are hard to move and compare;

cross-institution data comparison is impeded, due to installation-specific data model differences

LA tool adoption remains relatively low.

“analytics need to be broad-based, multi-sourced, contextual and

integrated” Siemens et al. [14],

7/15

Type of data gatheredType of data gathered

Analytic Tool Platform Data for analysis

SNAPP External tool Forum activity

LOCO-Analyst External tool Resource views, resource contents, forum contents

Course Signals LMSStudent age, residency, credits attempted, academic

history, course grades to date, interactions with the LMS

Desire2Learn Students Success System

LMSStudent grades, login frequency, discussion posts, results

and number of quiz attempts.

Open Learning Initiative MOOCs Knowledge Components achieved and failed

Khan Academic MOOCs Performance in exercises

Student-performed actions with a given outcome

Student Profile: age,

interests, etc.

8/15

IEEE LTSC 1484.11.1IEEE LTSC 1484.11.1

A.K.A. CMI data model (SCORM) “Bag” of records and fields

Student degree of progress “End State” (cmi.completion_status)

“State of Success” (cmi.success_status)

“Overall student performance” (cmi.score.raw)

Objectives degree of completion and success. progress

measurement, score…

9/15

IEEE LTSC 1484.11.1IEEE LTSC 1484.11.1

Interactions Store fine-grained information

Different kinds of interaction (relationship, true-false, etc. )

Multiple correct answers

Tagging particular entries with identifiers that link them to sets of related learning objectives

Journalist vs State

10/15

IEEE LTSC 1484.11.2IEEE LTSC 1484.11.2

API

11/15

Experience APIExperience API

Data Model Like Activity Stream

Extended for educational purposes

High flexibility

Verbs: key elements

<I> <did> <this>

<actor> <verb> <object>, with <result>, in <context>

12/15

LRS and APILRS and API

Learning Record Store

API

13/15

E-Learning standards + Learning AnalyticsE-Learning standards + Learning Analytics

Reduce development costs, protect investments…

Data reuse and broadens the pool of data that can be analyzed and explored

Move to other issues….

Requirements: Data model structure

Access and share data among systems

14/15

1484.111484.11

15/15

Experience APIExperience API

16/15

GraciasGracias

Ángel del Blanco Aguado

[email protected]

More information http://e-adventure.e-ucm.es

http://www.e-ucm.es

Publications http://www.e-ucm.es/publications/

17/15