Client: Media Lab Asia (Media Lab MIT)

ACM CHI 2003, Ft. Lauderdale, best paper award in the 2nd ACM CUU 2003,

Project goals

• Exploring various interaction paradigms for illiterate users through a collaborative method by including the potential users in the design process
• To build a simple, easy to use accounting application that aids the transactions within the small village-level Self Help groups.
• Building empowering tools for the community to facilitate the local economic processes of saving, borrowing, investing and earning.

Methodology

Phase 1
Study the existing work practices among the SHG group members in 4 villages. to construct an observation based model of user behavior with a focus on ‘the intuitive learning patterns of the semi-literate users to overcome the literacy barrier’
cognitive tests

Phase 2
Used task based test protocols and detailed tests for the conceptual design alternatives.

Tool Kits

• Photographs
• Open Interviews
• Wizard of oz testing
• Think aloud protocols
• Rapid onsite collaborative

Analysis

Phase 1
Qualitative analysis of data gathered during open interviews. Affinity diagramming to categories findings.
Quantitative and qualitative analysis of cognitive test results.

Phase 2
Data from ‘test’ sessions analysed for patterns in doing tasks and recording rate of task success with the alternative designs.

Synthesis

The iteration of the initial idea of numeric model of interaction revealed that it is difficult for these users to interpret numeric system as an abstraction for information hierarchy.
Further to this, the initial concepts built around a desktop computing paradigm faced many challenges, as the remote village locations do not have the basic provisions for installation for desktop systems.
We looked at the analyzed data to arrive at the most appropriate mode of computing vis-à-vis a sustainable form of interface technology among the given set of users
Therefore the Idea of a low cost data collection terminal as an effective and optimized mean of collecting transaction data was further explored.

Top