the idea:
While TouchViz was only one of several projects I worked on while at Microsoft Research, it was one of the most enjoyable. Working under Steven Drucker, a principal researcher at Microsoft Research, I was able to pour my energies into answering a question: "How would a user enact functions on a data visualization by direct, tactile manipulation of the visualization, instead of through selection of menus?"
The approach:
Because this was research, not product development, and I had neither users to test it on nor clear directives of what kind of data to manipulate, this project was, in terms of architecture, like building a dream house.There were pros and cons to the fact that it would never be put to the stress of actual users (expect within the confines of research purposes); It allowed me to freely explore what could be possible, but it also meant I wouldn't get the chance to re-iterate and improve the design based on user feedback. It was in the process of this project that I found both the joy of design research and its limitations.
Shake it up. Open brainstorming was the first step in the research process. Assuming that the visualization was on a handheld, but still fairly large product like an ipad, I started creating tables. First was a table of possible functions: if I wanted to order a bar graph from least-to-most, what was every possible way I could think of to do it? I was encouraged to think as out of the box as possible, with solutions ranging from tilting, shaking, drawing and dragging, pinching, spinning and flicking visualization shapes. Secondly, we arranged them by interaction types instead of function. From this table we then selected what were, to us, the most logical of these interaction types to fulfill each functional role. |
|
Lay it out.
A thorough storyboard was the next logical step. As we had no tools and no users at this stage (a prototype would be built later), I was forced to put it through its hypothetical paces with one of the most detailed storyboards of my career so far. I put together a user story for how they might want to explore a dataset, and then I started putting vector pen to paper, so to speak.
A thorough storyboard was the next logical step. As we had no tools and no users at this stage (a prototype would be built later), I was forced to put it through its hypothetical paces with one of the most detailed storyboards of my career so far. I put together a user story for how they might want to explore a dataset, and then I started putting vector pen to paper, so to speak.
Make it move.
While the storyboard was useful for a first test of the ideas in practice, there was a crucial aspect of the system that was lacking, in my opinion: the motion design. When the whole idea of the project is "tactile manipulation of data viz (shapes)", the way the shapes reacted to the manipulation seemed just as important as the manipulation method themselves. This video not only illustrates the delight of potential motion design in this context, but it served as a vision, an inspiration for the project. |
|
Test it. Write about it.
A rough prototype was put together by a developer working with Steven Drucker and I, and tested on users for viability as a design approach. A paper was published by ACM Sig CHI in 2013 titled TouchViz: A Case Study Comparing Two Interfaces for Data Analytics. |
All work copyrighted Microsoft Research unless otherwise noted.