Model for Rapid, Large-Scale Development of Learning Objects in Multiple Domains IEEE International Conference on Technology for Education (T4E) IIT Madras, Chennai July 14-16, 2011 --- Gargi Banerjee* Sahana Murthy* *Interdisciplinary Program in Educational Technology, IIT Bombay, Powai. Introduction Problem Statement Solutions Results 11Slide Conclusion e1 What are Learning Objects (LOs)? • Operational definition adopted – L’Allier’s Definition, since it has education as its focus. • Definition :“The smallest, independent structural experience that contains an objective, a learning activity and an assessment.” [L’Allier,1997] Sample LO Introduction Problem Statement Solutions Results Conclusion 22Slide e2 What are Learning Objects (LOs)? • Operational definition adopted – L’Allier’s Definition, since it has education as its focus. • Definition :“The smallest, independent structural experience that contains an objective, a learning activity and an assessment.” [L’Allier,1997] Sample LO Introduction Problem Statement Solutions Results Conclusion 33Slide e3 What are Learning Objects (LOs)? • Operational definition adopted – L’Allier’s Definition, since it has education as its focus. • Definition :“The smallest, independent structural experience that contains an objective, a learning activity and an assessment.” [L’Allier,1997] Sample LO Introduction Problem Statement Solutions Results Conclusion 44Slide e4 What are Learning Objects (LOs)? • Operational definition adopted – L’Allier’s Definition, since it has education as its focus. • Definition :“The smallest, independent structural experience that contains an objective, a learning activity and an assessment.” [L’Allier,1997] Sample LO Introduction Problem Statement Solutions Results Conclusion 55Slide e5 What are Learning Objects (LOs)? • Operational definition adopted – L’Allier’s Definition, since it has education as its focus. • Definition :“The smallest, independent structural experience that contains an objective, a learning activity and an assessment.” [L’Allier,1997] Sample LO Introduction Problem Statement Solutions Results Conclusion 66Slide e6 Why Learning Objects (LOs?) • Learning objects (LO) have become valuable educational resource a. Fit into different instructional settings b. Ideal for spreading higher education in India •Learning Object repositories (LORs) have increased rapidly in recent times Introduction Problem Statement Solutions Results Conclusion 77Slide e7 GOALS OF PROJECT OSCAR (IIT Bombay) • To create a repository of good quality open-source courseware of LOs • Educational Level targeted – Tertiary Level • Multiple Domains – Varied domains like Bioscience, Civil Engg., Chemistry, Earth Sc., Electrical Engg., Physics etc. • High production rate – 100 LOs in 1 year • Total of 300 LOs to be produced in 3 years • A ‘National Mission on Education through ICT’ (NMEICT) initiative • Funded by Ministry of Human Resources, GOI Introduction Problem Statement Solutions Results Conclusion 88Slide e8 How are LOs Produced ? Pedagogy expert SME Interface Designer Coders/Animators TYPICAL LO PRODUCTION PROCESS • Synchronous models studied – TEAL*1,MIDOA*2, PhET*3 •Good quality LOs produced but in restricted domains • Production Rate - 100 LOs in 4-5 years (approx.) 1 J. Belcher, “Studio Physics at MIT”. 2001, Available at: http://web.mit.edu/physics/news/physicsatmit/physicsatmit_01_teal.pdf 2 S. Arturo Barajas, A. Jaime Muñoz, R. Francisco J Álvarez., and G. Alma R.García, “Developing Large Scale Learning Objects for Software Engineering Process Model through MIDOA Model”, Mexican International Conference on Computer Science, September 2009 3 W. K Adams., S. Reid, R. LeMaster, S.B. McKagan, K.K. Perkins, C.E. Wieman, “A study of educational simulations Part I – Engagement and Learning”, Journal of Interactive Learning Research., 19(3). 397-419, 2008 Introduction Problem Statement Solutions Results 99Slide Conclusion e9 Problems faced in adopting a Synchronous Model • Scaling along numbers -Working simultaneously in multiple domains Team size increased Team geographically dispersed • Scaling along domains -a. Finding faculty in each domain willing to devote so much time b. Developers’ lack of domain knowledge aggravated ‘distance in communication’ problem Synchronous discussion hampered Conclusion : • Synchronous model will not work, if LOs are to be developed: a. in parallel in multiple domains b. when there is a time constraint to production of good quality LOs c. when a large development team is involved Introduction Problem Statement Solutions Results 1010Slide Conclusion e10 Broad Research Questions 1. What process to be adopted to create LOs in multiple domains at the tertiary level of education while maintaining high quality and high production rate? 2. Can an asynchronous, distributed development model replicate the efficiency of synchronous development model? Introduction Problem Statement Solutions Results 1111Slide Conclusion e11 Production Process Adopted Propose concept Review proposal • Concept Proposer Write IDD ppt • Reviewer Review IDD Stage 1 Introduction • IDD writer Develop Design • Reviewer Review Design Stage 2 Problem Statement • Designer • Reviewer Stage 3 Solutions Results Develop animation • Developer Review animation • Reviewer Stage 4 1212Slide Conclusion e12 Production Process Adopted • Concept Proposer Propose concept Review proposal Write IDD ppt • Reviewer Review IDD Stage 1 • • IDD writer Develop Design • Reviewer Review Design Stage 2 • Designer • Reviewer Stage 3 Develop animation • Developer Review animation • Reviewer Stage 4 Various asynchronous models of LO development tried A. Online Distributed Model C. Fully Outsourced Model B. Individual Faculty Model D. Domain Owner • In each of the models tried, the variable changed was : who was responsible for a specific task Introduction Problem Statement Solutions Results 1313Slide Conclusion e13 A. Online Distributed Model Process – Propose concept Review proposal • Concept Proposer = Online community • Reviewer = OSCAR Write IDD ppt Review IDD Stage 1 • IDD writer = Online community • Reviewer = IITB Faculty Stage 2 Develop Design Review Design • Designer = OSCAR • Reviewer = OSCAR Visual Design Expert Stage 3 Develop animation • Developer = OSCAR Review animation • Reviewer = IITB Faculty Stage 4 Advantages – • LOs developed by a community of teachers and learners. Limitations – • Content quality of the IDD biggest problem in this model. • Challenging to maintain communication channels between the developer, reviewer and the IDD writer (the person who originally proposed the idea) Lesson Learnt • Only if developer knew the domain could LOs under this model get developed Introduction Problem Statement Solutions Results 1414Slide Conclusion e14 B. Individual Faculty Model Process – • Concept Proposer = IITB Faculty Propose concept Write IDD ppt Review IDD Stage 1 • IDD writer = IITB Student Develop Design • Designer = OSCAR • Reviewer = IITB Faculty Review Design • Reviewer = OSCAR Visual Design Expert Stage 2 Develop animation • Developer = OSCAR Review animation • Reviewer = IITB Faculty Stage 3 Stage 4 Advantages – • Content quality of LOs ensured right from the beginning. Limitations – • • • Difficult to map a set of LOs to a course Content review took a very long time due to faculty time constraint Communication gap between the faculty SME and the developer Lesson Learnt: • Importance of standardizing instructions in asynchronous mode to replicate effectiveness of face-to-face interactions. Introduction Problem Statement Solutions Results 1515Slide Conclusion e15 C. Fully Outsourced Model Process – Propose concept Review proposal • Concept Proposer = IITB Faculty • Reviewer = OSCAR Write IDD ppt Review IDD Stage 1 • IDD writer =Vendor SME • Reviewer = IITB Faculty Stage 2 Develop Design Review Design • Designer = OSCAR • Reviewer = OSCAR Visual Design Expert Develop animation • Developer = Vendor Review animation • Reviewer = IITB Faculty Stage 3 Stage 4 Advantages – • Production cost low • LO coding time reduced Limitations – • LOs dispersed over various content areas, not coursewise • Gap in expected content standards between IITB SME and vendor SME. • Overall production time increased • Review process of IDD very lengthy • Vendor was constantly required to adapt to different expectations. Lessons Learnt - Content writing cannot be outsourced Introduction Problem Statement Solutions Results 1616Slide Conclusion e16 D. Domain Owner Model Process – Propose concept • Concept Proposer = IITB Faculty Review proposal Write IDD ppt Review IDD Stage 1 • IDD writer = IITB Student • Reviewer = IITB Faculty Develop Design Review Design Stage 2 • Designer = OSCAR • Reviewer = OSCAR Visual Design Expert Stage 3 Develop animation • Developer = Vendor Review animation • Reviewer = IITB Faculty Stage 4 Advantages – • Review time for both IDD and LO reduced • Instructional Designing training given to ID writers. Limitations – • Time lag between completion of the content generation phase and start of the LO coding phase Introduction Problem Statement Solutions Results 1717Slide Conclusion e17 D. Detailed Domain Owner Process SME Validate concept proposal Concept Proposal Stage Review of concept proposal OSCAR Instructional Design Document Validate requirement Validate pedagogy IITB students Pedagogy Instructional Designer Detailed Design Verify design Designer Pedagogy review IITB Faculty Pedagogy expert + IDD writer Design Review Design Reviewer Validate LO LO Development Vendor Developer IDD Review LO Review IITB Faculty Final LO Released Index: SME = Subject Matter expert, can be a faculty or a post-graduate/doctoral student; SB= Storyboard/IDD; LO = Learning Object Introduction Problem Statement Solutions Results 1818Slide Conclusion e18 D. Domain Owner Features: 1. Domain Owner model proved most successful within the constraints of cost, schedule and quality 2. Content creation done in-house, by faculty and students at our Institute 3. IDD development outsourced to animation companies 4. Vendor did not have to contend with different SME teams 5. Mean production time fell to 0.375 months for each LO without affecting quality [Table 1] 1919Slide e19 Results: Comparative Analysis Of Different Asynchronous Model Table 1: Development Data for all the models Model No. of IDDs submitted per month No. of approve d IDDs submitte d/month Mean Time taken for creating 1 approve d IDD [month] Mean Time taken to animate 1 approved IDD [month] Mean Time taken for 1 entire LO = approved IDD + LO development [month] Mean Cost of LO creation (Rs.) / LO Online 0.3 0 ∞ - - Individual Faculty 2 2 0.5 3mos/6LOs = 0.5mos 1 months 61,358/- Fully Outsourced 6 1 1 4mos/20LOs = 0.2mo 1.2months 24,600/- Domain Owner 4 4 0.25 2mos/16LOs = 0.125mos 0.375months 35,530/- - Index: mos = months Introduction Problem Statement Solutions Results 2020Slide Conclusion e20 Results: Comparative Analysis Of Different Asynchronous Model Introduction Problem Statement Solutions Results 2121Slide Conclusion e21 CONCLUSION • Domain Owner model seeks to harness the expertise of both academia and industry to create quality LOs • Content creation part of the LO production life-cycle is the most crucial •Social Impact of this model is enormous – a. OSCAR LOR is able to make LOs from advanced Science and Engineering domains accessible to students in every corner of India. b. Students and research scholars of our Institute, many of whom are faculty of the future, become aware of pedagogical issues Introduction Problem Statement Solutions Results 2222Slide Conclusion e22 ACKNOWLEDGEMENT • The authors would like to thank the following organizations and people for their contribution, help and support: NMEICT , Ministry of Human Resources & Development, Govt. of India IIT Bombay Project OSCAR team : PIs, Project Managers, ID writers, Designers, Animators IIT Bombay faculty All Domain Owners in Project OSCAR 2323Slide e23 THANK YOU ! 2424Slide e24