MOOCs and online learning:learner support and teacher workload

Diana LaurillardLondon Knowledge LabInstitute of Education

08 July 2013

The issues

• Global demand for HE

• The aims of HE

• The roles of TEL

• Modelling MOOCs

• Teachers as innovators

The Challenges to Higher Education

The global demand for HE

The new UNESCO goals for education:• Every child completes a full 9 years of free

basic education … • Post-basic education expanded to meet needs

for knowledge and skills … (Draft for UNESCO post 2015 goals)

By 2025, the global demand for higher education will double to ~200m per year, mostly from emerging economies (NAFSA 2010)

Implying significant graduate and teacher training needs for this level of schooling and HE 1:25 staff:students??

“The knowledge industry is regarded as a global business… Because HE is seen as a ‘market’ game rather than a social project… universities no longer have to apologize for treating international students as ‘customers’ who represent a significant income stream” (Scott, 2011: 64)

UK Export income of HEIs: £2.4bn fees; £4.3bn for living

So online teaching as an export generates much less ROI

We are now ‘entrepreneurial’ universities, but which of our ‘classic’ values do we preserve?

Universities as businesses

The aims and purpose of HE

The purposes of higher education:• To inspire and enable individuals to develop their

capabilities to the highest…• To increase knowledge and understanding for their

own sake…• To serve the needs of an adaptable, sustainable

knowledge-based economy…• To play a major role in shaping a democratic,

civilised, inclusive society (NCIHE, 1997)

How could mass HE achieve that nurturing of the individual, while reducing the current 1:25 staff:student ratio for student support?

The fundamental academic aims

To improve the quality and scale of HE across the world

- Not to access masses of data from desperate would-be students- Not simply to market our existing courses more widely- Not to provide free CPD to employed professionals

• Adaptive feedback (simulation/models)

• Expositions (lecture videos)

• Automated grading (MCQs, quizzes)

• Readings (pdfs)

• Collaboration activities (wiki)

• Peer group discussion (forums)

• Peer grading against criteria (tests)

• Tutored discussion (forums)

• Tutor feedback (e-portfolio)

• Adaptive feedback (simulation/models)

• Expositions (lecture videos)

• Automated grading (MCQs, quizzes)

• Readings (pdfs)

• Collaboration activities (wiki)

• Peer group discussion (forums)

• Peer grading against criteria (tests)

• Tutored discussion (forums)

• Tutor feedback (e-portfolio)

• Adaptive feedback (simulation/models)

• Expositions (lecture videos)

• Automated grading (MCQs, quizzes)

• Readings (pdfs)

• Collaboration activities (wiki)

• Peer group discussion (forums)

• Peer grading against criteria (tests)

• Tutored discussion (forums)

• Tutor feedback (e-portfolio)

Understanding high quality T&L

MOOC vs standard online course Preparation time (fixed costs)

Support time (variable costs)

Lessons from Georgia Tech MOOC

IssuesAT&T is donating $2-million; Bill & Melinda Gates Foundation supportSome students want a great deal of guidanceEvaluating students’ writing can only rely on peer reviewPlatform constraints affect pedagogy:

Not allowed to a fail a student who gives no feedbackSo many students get no peer feedbackStudents can automatically peer review—even if they failed the training

Good pointsStudents form communities within the discussion forumsPeer help with questions about content or technologyMore ambitious students have developed study guidesSome self-identified instructors have formed their own forum, to consider how they can use our course to teach their own students

The Duke MOOC

Bioelectricity: A Quantitative Approach Taught in class for over 20 yearsExperimental move to a free and open MOOC12,000 students enrolled from >100 countries• 8 weeks long • 97 ~6 min videos • 22 GB of data • 1052 files • 18 graded exercises, including a peer-graded writing

assignment and final exam (Duke University 2013)

The Duke MOOC

Not for undergraduates

Potential undergraduates

Enrolled students

The Edinburgh MOOCs

Not for undergraduates

Enrolled students

Less than high school

School

College

Degree

PG degree

0% 5% 10%15%20%25%30%35%40%45%

Potential undergraduates

40%

30%

17%

10%

3%

MOOCs @ Edinburgh 2013 – Report #1

The Duke MOOC

Not for the faint-hearted

Comparable with normal online u/g

courses = 2% of enrolment

The Edinburgh MOOCs

Average student numbers per course

Statement of Accomplishment

Week 5 asst's

Engaged Week 1

Accessed Week 1

Enrolled

0 10000 20000 30000 40000 50000 60000

5500

6000

15000

20500

51500

= 10% of enrolment

420 hours to develop materials and course design

Basic MOOCs vs the Duke MOOC: Comparing the learning experience

Videos and pdfsQuizzesWikiPeer discussionsPeer gradingTutored discussionsSummative assessment

High on prep timeZero contact for 42 hours

Low on prep timeHigh contact for 8 hours

200 hours to support ~500 students for 8 hours = 1:20 staff student ratio

Basic: 8 weeks, providing 50 hours learning time, no support

Report at http://bit.ly/ZRMbjp

Duke: 8 weeks, providing 50 hours learning time:

Comparing teacher hours for a basic MOOC and the Duke MOOC (48 hour course)

Support time 50 500 5000Duke MOOC 20 hrs 200 hrs 2000 hrsBasic MOOC 0.00 0.00 0.00

Teacher support time rises to 2000 hours for 5000 students.

2000 hours = 1 year of a tutor for a 5 credit course.

= 24 FT tutors for 120 credit course.

50 500 50000

500

1000

1500

2000

2500

3000

Duke MOOCBasic MOOC

50 500 50000

500

1000

1500

2000

2500

3000

Duke MOOCBasic MOOC

Total teacher time

Preparation time = 420 hrsThe variable cost of support does not achieve economies of scale

Prep time = 420

Modelling the benefits and costs

• We need to understand the pedagogical benefits and teacher time costs of online HE

• What are the new digital pedagogies that will address the 1:25 student support conundrum?

• How do we turn variable-cost pedagogies into fixed-cost?

Large-scale pedagogy (Edinburgh MOOCs)

Academic reads posts selectively and summarises each week so not a variable cost

+ Popular with students

Pedagogies for supporting large classes

Concealed MCQs

The virtual Keller Plan

The vicarious master class

Pyramid discussion groups

Tutorial for 5 representative studentsQuestions and guidance represent all students’ needs

Conceal answers to questionAsk for user-constructed input Reveal multiple answersAsk user to select nearest fit

240 individual students produce response to open questionPairs compare and produce joint responseGroups of 4 compare and produce joint response and post as one of 10 responses...6 groups of 40 students vote on best responseTeacher receives 6 responses to comment on

Introduce contentSelf-paced practiceTutor-marked testStudent becomes tutor for creditUntil half class is tutoring the rest

Teachers as designers need the tools for innovation

Tools for teachers as learning designers

To find or create new ideasAdoptAdaptTest

To collect learning analyticsRedesignAnalysePublish

Creating knowledge about effective blended and online pedagogies

Laurillard, D. (2012). Teaching as a Design Science: Building Pedagogical Patterns for Learning and Technology. New York and London: Routledge.

Tools for teachers as learning designers

http://tinyurl.com/ppcollector

A library of patterns to

inspect

Teachers sharing their best designs

Colour-coded text identifies content

parametersBlack text expresses

pedagogy design

Capturing their mixed mode pedagogies

Category of • learning type• duration in minutes• teacher contact/not

Defining the metadata of their pedagogies

Export to Moodle for Ed students

Reversioned for Med students

The cycle of professional collaboration:Search - Adopt – Adapt – Link resources and tools –

Test – Revise - Annotate – Export to VLE – Publish to repository – Search

Turning Mode 2 knowledge about the practice of teaching into Mode 1 knowledge that is explicit,

articulated, testable, shareable, generalisable and localisable (Gibbons et al, 1994)

AcquisitionInquiryDiscussionPracticeProduction

Acquisition

Inquiry

Discussion

Practice

Production

Conventional

Blended

Categorised learning activities

Analysis shows more active learning

A computational representation can analyse how much of each activity has been designed in

Modelling the pedagogic benefits

Learner hours

Learning activities Group size

Weekly Non-weekly Total

Tutored discussion 15 3.3 40Readings 15 6.7 80Formative practice (tutor) 1 22.0 22Formative practice (peer) 1 0Formative practice (computer) 1 0Summative assessment (tutor) 1 23.0 23Building up own notes 15 3.3 40Exploring resources 15 2.1 25Application of concepts 15 5.0 60Personal tuition 1 1.5 1.5Self-directed learning 8.5 8.5Total learning hours 300

AcquisitionInquiryDiscussionPracticeProduction

PersonalisedSocialOne-size-for-all

Modelling the benefits of blended courses

Academics define the • mix of physical and

digital activities, • group size, and• distribution of total

learning hours

Figure 2(b) Teaching time for a course with 40, 80, 160 students, gives profits of -£12000 £13000 £35000

Figure 2(a) Teaching time for a course with 40 students each year, gives profits of -£12000 £5000 £8000

Modelling the costs of online courses

Modelling an IOE course over 3 years

Prep hrs

Support hrs

Yr1 Yr2 Yr3

Prep hrs

Support hrs

Yr1 Yr2 Yr3

Modelling the costs for increasing student cohort size

Scaling up to large numbers will never improve the per-student support costs…

…unless we come up with some clever pedagogical patterns that support at better than the 1:25 ratio

The question is – what are they, and how do we develop and share them?

An institutional approach to blended learning

Create a ‘T&L’ learning organisation:

• Changes to T&L are modelled carefully

• Model the University’s preferred pedagogies

• Innovation is designed to collect and use evidence

Invest in teachers innovating

Teaching innovation is rewarded alongside research

Reconceptualise teaching as knowledge building

Start from the vision / problem, not the technology

The Senior Team must ask ‘how can technology help?’

Teaching as a Design Science: Building pedagogical patterns for learning and technology (Routledge, 2012)

d.laurillard@ioe.ac.uk

www.ldse.org.uk/

tinyurl.com/ppcollector

Further details…

The ALT MOOC ‘OCTEL’ Open Course in Technology Enhanced Learningat http://octel.alt.ac.uk/ April 2013

The global demand for HE requires investment in pedagogic innovation for MOOCs to deliver

TEL-based pedagogic innovation must support students at a better than 1:25 staff-student ratio

Teachers need the tools to design, test, gather the evidence of what works, model benefits and costs

Teachers are the engine of innovation – designing, testing, sharing their best pedagogic ideas

Teaching as a design science: Tools for professionals teaching MOOCs

The global demand for HE requires investment in pedagogic innovation for MOOCs to deliver

TEL-based pedagogic innovation must support students at a better than 1:25 staff-student ratio

Teachers need the tools to design, test, gather the evidence of what works, model benefits and costs

Teachers are the engine of innovation – designing, testing, sharing their best pedagogic ideas