social learning at massive scale lwmoocs 2015 slideshare

Mike SharplesTwitter: @sharplm

The Open University

Social Learning at Massive Scale

Massive Open Online Courses FUTURELEARN

193 courses from 72 partnersBroad range of

business, health, science, arts

Male:41%

Female:59%

Responsive platform

25% access on mobile devices

Over 2 million learners

4 million course sign-ups

Company formed by The Open University, launched in October 2013

29 of the UK’s leading universitiesFUTURELEARN PARTNERS

PAGE 4

21 Global UniversitiesFUTURELEARN PARTNERS

22 Specialist OrganisationsFUTURELEARN PARTNERS

The BBC has co-developed its first four courses with partner universitiesA BROADER RANGE OF CULTURAL AND CONTENT PARTNERSHIPS

BBC

440,000 registrations (270,00 learners) for Understanding IELTS course from the British Council

WORLD’S BIGGEST MOOC

‘Institutional MOOC strategies in Europe’ - EADTUDIFFERENCES BETWEEN US AND EUROPE

EATDUhttp://www.eadtu.eu/documents/Publications/OEenM/Institutional_MOOC_strategies_in_Europe.pdf

PEDAGOGY-INFORMED DESIGN

How do we design successful pedagogy-informed massive open courses?

PEDAGOGY OF MOOCS

http://en.m.wikipedia.org/wiki/File:Figure_1_MOOCs_and_Open_Education_Timeline_p6.jpg Creative Commons Attribution 3.0 Unported license.

Connectivist learning

Instructivist learning

Social constructivist learning

Open learningSocial constructivism engages learners in working together to

construct and share knowledge

http://en.m.wikipedia.org/wiki/File:Figure_1_MOOCs_and_Open_Education_Timeline_p6.jpg

http://en.wikipedia.org/wiki/en:Creative_Commons

http://creativecommons.org/licenses/by/3.0/deed.en

PEDAGOGY-INFORMED DESIGNSocial learning

A.N. Meltzoff, P. K. Kuhl, J. Movellan, & T. J. Sejnowski (200) Foundations for a New Science of Learning, Science 325 (5938), 284.

“A key component is the role of ‘the social’ in learning. What makes social interaction such a powerful catalyst for learning?”




John Hattie – Visible Learning

“What is most important is that teaching is visible to the student, and that the learning is visible to the teacher. The more the student becomes the teacher, and the more the teacher becomes the learner, then the more successful are the outcomes ” Hattie, p.17

PEDAGOGY-INFORMED DESIGNStorytelling and narrative

“Narrative is central to our cognition … It helps us think, remember, communicate, and make sense of the world… Learners constantly adjust their understanding in accordance with their exposure to conventional narratives, making the construction of narrative a central cognitive goal.”Plowman et al. (1999) Designing multimedia for learning: Narrative guidance and narrative construction. CHI ’99.


Page


FORMATIVE ASSESSMENT AND MASTERY LEARNING

COLLABORATIVE LEARNING

REVIEW AND FEEDBACK

REPUTATION AND REWARD

VISIBLE LEARNING AND GOAL SETTING

NARRATIVE LEARNING AND STORYTELLING

We are not simply adopting methods from previous platforms such as Moodle. We have to rethink• for massive scale• a wide variety of learners

with different abilities• self-motivation rather than

learning for credit• multiple devices

Effective methods of teaching, learning and assessment

SCALING LEARNING

Some educational methods get worse with scale

e.g. personal tutoring, sports coaching

Some educational methods don’t change with scale

e.g. lecturing, testingWhich educational methods

improve with scale?

Metcalfe, R.M.: It’s all in your head. Forbes, 20th April, 2007. http://www.forbes.com/forbes/2007/0507/052.html

METCALFE’S LAW

For some networked systemsthe value of a product or service increases with the number of people using it

Telephone network

Downes, S.: The Personal Network Effect. Blog posting, 4th November 2007. http://halfanhour.blogspot.co.uk/2007/11/personal-network-effect.html.

LEARNING NETWORKS

Networks enable learning if they support conversations that are

new, important, timely, usable, understandable, appropriate, trusted

Develop effective social networks for learning

TWO-SIDED NETWORKS

Opportunity to teach

Manage complexity

Gain insight from data

Opportunity to learn

Ease of use

Value that increases with scale

For a 2-sided network, both sides need to benefit

EducatorsLearners

Learning as conversation (Pask, Laurillard)MASSIVE SCALE SOCIAL LEARNING

Learner • demonstrates understanding

• proposes solutions to problems

Learner • acts to develop models• acts to solve problems

Partner • demonstrates understanding

• elaborates solutions to problems

Partner • acts to develop models • helps to solve problems

Level of descriptions

Level of actions

Shared medium • enables learners and partners to represent arguments and reach

agreements

Shared medium • enables learners and partners to access

information, develop models and solve problems

‘what’ & ‘how?’ questions and responses

‘why?’ questions and responses

offering conceptions and explanations

proposing goals and modifying actions

reflect

adapt

reflect

adapt

MASSIVE SCALE SOCIAL LEARNING

Enabling effective conversations for learningAt levels of action (about activities) and description (about concepts)For hundreds of thousands of peopleFrom many cultures

Conversations for learning

A visible and easy to follow structure and progressVISIBLE LEARNING

Courses are made up of 6-8 weeksWeeks run Mon-Sun and have a title

telling you what you’re going to learn

Weeks are made up of activitiesAn activity is a set of steps with

a clear learning objective

Activities are made up of stepsA step could be an article, video,

discussion, test, quiz, assignment…

Practical activity linked to conversationLEARNING THROUGH ACTIVITY

270,000 learners

From around the world

For many, their first time learning online

35% contributed to discussions

25% on mobile devices

Apply methods from social networking to manage the massive-scale conversations

SOCIAL NETWORK METHODS

Replies Likes FilteringFollowing




Partner• demonstrates understanding


Partner• acts to develop models • helps to solve problems


Level of actions

Shared medium • enables learners and partners to represent

arguments and reach agreements



‘how?’ questions and responses




reflect

adapt

reflect

adaptDiscussions linked to content

Peer review

Exercise steps

Insightful Analytics from learning at massive scale

DEMOGRAPHIC PROFILEFutureLearn is attracting a high proportion of female learners and an even spread of age groups

61% Female

39% Male Age in

Years

Under 18 years

old

18 - 25 years

old

26 - 35 years

old

36 - 45 years

old

46 - 55 years

old

56 - 65 years

old

Over 66 years

old

0%

5%

10%

15%

20%

25%

Data from Pre-Course Survey January 2015

40%

2.6m JoinersSomeone who signed up for the course.

84%

ActiveLearnersLearners who marked steps as complete

43%ReturningLearnersLearners who marked steps as complete in more than one week

23%Fully Participating LearnersLearners who marked the majority of steps as complete including all of the assessments

Learners who make comments

52%Learners

Joiners who viewed at least one step of the

course

15

Data provided by FutureLearn from all 51 courses starting between 29 July 2013 and 1 December 2014, lasting 5 weeks or more, and offering tests.

Coursera: N. Gillani, R. Eynon, Communication Patterns in Massively Open Online Courses. The Internet and Higher Ed 23, 18-26 (2014).

EdX: A.D. Ho et al., HarvardX and MITx: “The first Year of Open Online Courses” (Working paper no. 1, HarvardX and MITx, Cambridge, MA, 2014).

Edinburgh Coursera data provided by University of Edinburgh for 22 Coursera courses to 1st March 2015

COMPARISON WITH OTHER PLATFORMSFutureLearn data 2013-2014 for courses of 5 weeks or longer, with tests

The importance of responsive designANYTIME LEARNING

Learning for people’s lives

While desktop users still make up the majority of our learners, tablet usage increases in the evenings, and mobile access remains constant through the day.

Mon Tues Wed SatThur Fri Sun12pm 12pm 12pm 12pm 12pm 12pm 12pm

Time

Visit

s

Desktop Tablet Mobile

Video duration vs session exit rateANALYTICS FOR LEARNING DESIGN

Short videos retain learners

Our goal is for steps to take no longer than 20 minutes.When videos last beyond 10 minutes, learners start to exit the site.

Video Length

Exit

%

0 10 20 30

40

60

20

80

Duration of peer review assignment vs length of assignmentANALYTICS FOR PLATFORM DEVELOPMENT

Looking for patterns

Most assignments are around the guide length of 200 words.Some assignments are very short (0-10 words) and can be ignored for review.Some assignments take less than a minute between first view and submission, but are up to 1000 words – suggests these learners copy and paste.

I am writing to you from the Medecins Sans Frontieres Bo-Ebola Treatment Centre in Sierra Leone. As an expat doctor, I have found your Course on "Ebola In Context" from the London School of Tropical Medicine and Hygiene to be excellent. Our National Staff, who are local Sierra Leone Nurses and Clinical Officers have enrolled in the course on their mobile phones, however because internet is poor or not available, they have been unable to attend the course/stream videos or download. In turn, with the help of MSF, I have been able to download the videos, and articles, and together in a group of around 40 people, we have completed your course.

Courses make a real difference LEARNER FEEDBACK

Sincerely, Keri Cohn, MD, MPH, DTMH Expatriate Responsible Doctor

Bo-Ebola Treatment Center, Sierra Leone Medecins Sans Frontieres- OCA- Amsterdam





Educator• demonstrates understanding


Educator• acts to develop models • helps to solve problems


Level of actions


agreements







reflect

adapt

reflect


Peer review

Analytics for learning design

Exercise steps





Partner• demonstrates understanding


Partner• acts to develop models • helps to solve problems


Level of actions


agreements







reflect

adapt

reflect


Peer review

Analytics for learning designSocial annotation

Group learning Peer assessment

Exercise steps

ArgueGraph

MOOCS

How do we design successful open learning at massive scale?

Massive scale open courses

Pedagogies that improve with scale

Connected with insightful analytics

Learning for life