Upload
cooperacion-20-2009
View
1.067
Download
2
Tags:
Embed Size (px)
DESCRIPTION
Kentaro Toyama, Investigando en TIC para el Desarrollo Humano, para el II Encuentro Internacional TIC para la Cooperación al Desarrollo.
Citation preview
Research for Developmentat Microsoft Research India
Kentaro ToyamaAssistant Managing Director
Microsoft Research India
Cooperación 2.0February 11, 2009 – Gijón
Outline
Introduction to Microsoft Research India
Methodology and Sample Projects
Key Lessons
Outline
Introduction to Microsoft Research India
Methodology and Sample Projects
Key Lessons
Microsoft Research IndiaEstablished January, 2005
Seven research areas– Algorithms– Cryptography, Security & Applied Math– Digital Geographics– Mobility, Networks, and Systems– Multilingual Systems– Rigorous Software Engineering– Technology for Emerging Markets
Contributions to Microsoft:– MultiPoint, Netra, Virtual India
Currently ~60 full-time staff, growing
Collaborations with government, academia, industry, and NGOs in India
Microsoft Research IndiaSadashivnagar, Bangalorehttp://research.microsoft.com/india
“Technology for Emerging Markets”
Understand potential technology users in developing communities
Design and evaluate technology and systems that contribute to socio-economic development of poor communities worldwide
Collaborate with development-focused organizations for sustained, scaled impact
Computer-skills camp in Nakalabande, Bangalore(MSR India, Stree Jagruti Samiti, St. Joseph’s College)
Research Group Goals
Multidisciplinary ResearchAishwarya Lakshmi Ratan
–Public Administration and International Development
Jonathan Donner
– Communications
Nimmi Rangaswamy
– Social Anthropology
Indrani Medhi– Design
Kentaro Toyama (Group Lead)
– Computer Science
Society
Group
Technology
Individual
Society
Group
Technology
Individual
Design
Understanding
Impa
ct
Understanding
Impa
ct
Rikin Gandhi– Astrophysics
Saurabh Panjwani– Computer Science
David Hutchful– Human Computer Interaction
Bill Thies– Computer Science
Microsoft Confidential
Research Sites
- Other projects studied
- Our projects
ICTD Conference
Co-founded by MSR India, UC Berkeley, MIT, CMU, IIIT-Bangalore
Focus on rigorous academic work, with all papers double-blind peer-reviewed
Established a multidisciplinary community of academic researchers in technology for development
First: May 25-26, 2006, Berkeley (UCB)
Second: Dec 15-16, 2007, Bangalore (MSR)
Third: April 17-19 2009, Doha, Qatar (CMU)
IEEE/ACM International Conference on Information and Communication Technologies and Development
UC Berkeley, site of ICTD 2006
Outline
Introduction to Microsoft Research India
Methodology and Sample Projects
Key Lessons
Our MethodologyImmersion
– Methodology: ethnography • qualitative social science
Design– Methodology: iterated prototyping
• design, engineering
Evaluation– Methodology: randomized control trial
• economics
Implementation– Methodology: partnership
• political science
( )
Par
tner
ship
NG
Os,
gov
ern
me
nts,
loca
l fir
ms,
com
mu
niti
es
4.114.56
3.73.76
2.93
4.53
3.6
2.8
4.44.3 4.54.1
0
1
2
3
4
5
ALL STUDENTS BOYS GIRLS
Aver
age
No.
of W
ords
Lear
nt
SS MS MM-R MM-V
SS
MS
MM
-R
MM
-V
Kelsa+
Microfinance & Technology
IT and Microentrepreneurs
Information ecology of very small businesses
Potential of technology to support microfinance
MicroenterprisePC + mobileMixed-method studyResearch only
Text-Free UI
Text-free user interfaces fornon-literate users
User interfacesPCDesignUser studiesGuidelines
MicrofinancePC + mobileQualitative studiesBusiness analysisResearch only
Sample Projects
Information accessPCQualitative studyUsage analysisPilot
Featherweight Multimedia
Paper and cheap electronicsfor low-cost multimedia
General educationElectronicsHCIUser studiesPrelim research
MultiPoint
Primary educationPCHCIUser studiesSoftware SDK
Warana Unwired
Digital Green
Video and mediated instructionfor agriculture extension
Substitution of mobile phones for rural PC kiosks
AgricultureVideoInterventionControl trialsPilot
Info systemsMobileInterventionRural kiosksPilot
Free access PCs for low-income office staff
Multi-user systems for computer labs in schools
Microfinance and
TechnologyMultiPoint Digital Green
Partnership
Pradan, Ujjivan, Sanghamitra, CCD, BASIX, Mahakalasam…
CLT, Azim Premji Foundation, local gov’t schools, etc.
GREEN Foundation, PRADAN, etc.
ImmersionAnalyze process flow and costs
Observe use of computers in rural schools
Understand farmers' needs and capabilities
Design None so far“One mouse per child”
Participatory video & mediated instruction
Evaluation238 students, against one PC per child
1 year, 20 villages, against classic extension
ImplementationCost analysis spreadsheet available online
New Microsoft product
Spin off independent NGO
Exploratory Studies Site visits:• Interviews with…
– Institution heads– MFI agents– Clients
• Participant observation• Accounts and records
Microfinance Institutions• Pradan• Ujjivan• Sanghamitra• CCD Mahakalasam• BASIX• Etc.
Microfinance and Technology
Work by Aishwarya Ratan
Classic banking, USA Microcredit, India
$22,000/yr $1200/yr $150*$1000*
10% increase in productivity
Is Technology Always Worth It?
$220010% increase in productivity $120
Cost of technology may outweigh benefit.
* Estimated amortized cost of device per annum including maintenance and support
Microfinance and Technology
Can technology assist microfinance?
Front-end IS1. Account creation (loan,
savings & insurance)1. Collecting client data2. Screening/ verification
2. Transaction data3. Processing claims (savings,
transfers & insurance)
E-paymentsEnabling e-cash transactions
1. Disbursal of amount (loan)2. Collection of dues/ payments (loan, savings &
insurance)
Back-end IS1. Aggregation of client
data1. Actuarial analysis2. Target offerings YES
MAYBE?
Difficult
Microfinance and Technology
Current Status
Costing spreadsheet available to assess relative value of technology intervention for MFIs
Aborted a mobile-phone tool for loan officers
Ethnographic study of mobile payment usage among low-income communities in four countries with CGAP
Ongoing research on optimal UI for non-literate users of mobile banking systems with CGAP
Ratan, Aishwarya. L. and Mahesh Gogineni. (2008, May). Cost Realism in Deploying Technologies for Development. Paper presented at the Conference on “Confronting the Challenge of Technology for Development: Experiences from the BRICS”, Department of International Development, University of Oxford, 29-30 May 2008.Ratan, Aishwarya. (2006, December). Deconstructing 4 Microfinance Myths. Microsoft Research Technical Report.
Microfinance and Technology
Microfinance and
TechnologyMultiPoint Digital Green
Partnership
Pradan, Ujjivan, Sanghamitra, CCD, BASIX, Mahakalasam…
CLT, Azim Premji Foundation, local gov’t schools, etc.
GREEN Foundation, PRADAN, etc.
ImmersionAnalyze process flow and costs
Observe use of computers in rural schools
Understand farmers' needs and capabilities
Design None so far“One mouse per child”
Participatory video & mediated instruction
Evaluation238 students, against one PC per child
1 year, 20 villages, against classic extension
ImplementationCost analysis spreadsheet available online
New Microsoft product
Spin off independent NGO
PCs in SchoolsSchools…
•Teachers under-prepared for computer skills•Financing for PC systems erratic
Parents…
•Happy to see PCs in schools; want children to learn. •Have little understanding of PC functionality•PC “mastery” believed by some to come quicker than English ability (though English ability more valued)
Strong anecdotal evidence that children attend school longer, if there are PCs at the school for student use.
If schools have PCs, their use is many students per PC.
MultiPoint
Work by Joyojeet Pal, Meera Lakshmanan, Kentaro Toyama
Invention
Provide a mouse for every student
– One cursor for each mouse, with different colours or shapes
– USB mice• Experimented with up to 20• (Theoretically works up to 128)
– Reduces per-student cost of interaction
– Content modified • Game-like environment
MultiPoint
Work by Udai Pawar, Kentaro Toyama, Rahul Gupta
Evaluation
After MultiPointBefore MultiPoint
MultiPoint
Average number of words learned during PC usage
4.114.56
3.73.76
2.93
4.53
3.6
2.8
4.44.3 4.54.1
0
1
2
3
4
5
ALL STUDENTS BOYS GIRLS
Aver
age
No.
of W
ords
Lear
nt
SS MS MM-R MM-V
SS
MS
MM
-R
MM
-V
Observations:
– Strong gender effects– Girls cooperate and learn
regardless of mode– Boys compete blindly without
cooperative structure
– Varying social engagement between students depending on mode
– Dominance– Collaboration
– Identification with cursor– MultiPoint with voting
appears most constructive
Number of words learned under MultiPoint roughly the same as with SS.
Current Status
Microsoft released free MultiPoint SDK, June 2007
Additional studies with MultiPoint around collaboration and creative software
Mouse Mischief: MultiPoint for the entire class (Neema Moraveji)
Split Screen studies continuing
New hypothesis: Better anywhere for primary education over one PC per child?
Pawar, U. S., Pal, J., and Toyama, K. (2006) Multiple mice for computers in education in developing countries, IEEE/ACM Int’l Conf. on Information & Communication Technologies for Development, ICTD 2006 .Pawar, U.S., Pal, J., Gupta. R., and Toyama, K. (2007) Multiple Mice for Retention Tasks in Disadvantaged Schools, In Proceedings of ACM CHI’07, ACM Press.
http://thescooterlounge.com/images/124IndianFamily.jpg
Sharing hardware not unusual in India
MultiPoint
Microfinance and
TechnologyMultiPoint Digital Green
Partnership
Pradan, Ujjivan, Sanghamitra, CCD, BASIX, Mahakalasam…
CLT, Azim Premji Foundation, local gov’t schools, etc.
GREEN Foundation, PRADAN, etc.
ImmersionAnalyze process flow and costs
Observe use of computers in rural schools
Understand farmers' needs and capabilities
Design None so far“One mouse per child”
Participatory video & mediated instruction
Evaluation238 students, against one PC per child
1 year, 20 villages, against classic extension
ImplementationCost analysis spreadsheet available online
New Microsoft product
Spin off independent NGO
Eight months over 2005 & 2007
• Participant observation• Structured interviews
– 200 farmers (users and non-users)
– 3 NGO leaders– 12 NGO staff
• Technical analysis• Surveys
Background knowledge of agriculture in several regions
Thorough knowledge of Indian agriculture extension
Ethnographic Investigation
Photo: Rajesh Veeraraghavan
Digital Green
Work by Rajesh Veeraraghavn, Rikin Gandhi
Digital Green
System Design
Participatory content production•Standard extension procedure•Rough “storyboarding”
– Repetitive pattern; easy to learn– Minimize post-production
•Local farmers on their own fields– Reduce perception of “teachers”– Promote “local stars”
Mediated instruction•Locally hired mediators
– Encourage discussion
•On-demand screenings – Choice time and place– Not “stand-alone” kiosk
Work by Rikin Gandhi, Vanaja Ramprasad
7 times more adoptions; 10 times more cost-effective, over classical extension
15 months: 13 villages, 3 nights a week, 1,000 regulars
Sustained local presence
Mediation
Repetition (and novelty)
Integration into existing extension operations
Social homophily between mediator, actor, and farmer
Desire to be “on TV”
Trust built from identities of farmers and villages in videos
Evaluation Results
Work by Rikin Gandhi, Rajesh Veeraraghavan, Kentaro Toyama
Participatory video and mediated instruction enables 10x increase in cost-effectiveness of traditional agriculture extension.
Gandhi, R., R. Veeraraghavan, K. Toyama, V. Ramprasad. Digital Green: Participatory Video for Agricultural Extension, in Proc. Annual Meetings of American Society of Agronomy, 2007.
Stockholm Challenge Award 2008
Now in 12 villages, impacting ~2000 households
Discussions with BAIF, PRADAN, others
Spinning off independent NGO to scale Digital Green:
Aiming to impact ~400,000 households in 3 years
Current Status
Microfinance and
TechnologyMultiPoint Digital Green
Partnership
Pradan, Ujjivan, Sanghamitra, CCD, BASIX, Mahakalasam…
CLT, Azim Premji Foundation, local gov’t schools, etc.
GREEN Foundation, PRADAN, etc.
ImmersionAnalyze process flow and costs
Observe use of computers in rural schools
Understand farmers' needs and capabilities
Design None so far“One mouse per child”
Participatory video & mediated instruction
Evaluation238 students, against one PC per child
1 year, 20 villages, against classic extension
ImplementationCost analysis spreadsheet available online
New Microsoft product
Spin off independent NGO
Outline
Introduction to Microsoft Research India
Methodology and Sample Projects
Key Lessons
Key Lessons
Development first (not technology first)
Time with communities (not with “experts”)
Multidisciplinarity of teams (not individuals)
Quality through great people (not processes)
Sustainability is case-by-case (not by magic formula)
Impact as the goal (not ideology)
Stronglyrecommended,but some successful counterexamplesexist.
Summary
Introduction to Microsoft Research India
Methodology Sample Projects– Immersion Microfinance & tech– Design MultiPoint– Evaluation Digital Green
Key Lessons
Thank you!http://research.microsoft.com/research/tem