Halloween makes today’s tea time a very special meeting. We have injured Greg (Michael), the fonz (Greg), Sarah Palin (Lydia), Theodore Robosevelt (Matthew) and track start (Katrina). All the rest people are in original dressing.
We are planning a class of interaction design during IAP 2009. This would be a project-oriented course that focuses on designing components and teaches how to brainstorming, sketching ideas , and prototyping. The class requires not only brainstorming but also “bodystorming”, which means you have to interact with your team members and get your ideas done. We may teach how to use some prototyping techniques, such as paper prototyping, video prototyping, and Flash or other web prototyping tools.
It would be a four-week long course, and would have 2 small projects and 1 final project. The two main tracks of it are prototyping and brainstorming. To fill the gap to the course 6.831, User Interface Design and Implementation, it would be designed as a supplement to 6.831. We hope every student can learn how to turn ideas into real stuff and get feedback quickly after this class.
The feed of this blog has been integrated to our UID web site!
What do you think a computer scientist can do in NASA? Aritificial Intelligence and robot navigation? Jeff Norris told us It was an old view about NASA!
In addition to AI planners for Mars exploration, they also need lots of HCI work too, to support their scientists and operations engineers authoring plans for the robots and viewing and analyzing lots of data and pictures. In project Maestro, a team in JPL uses an open source tool, Eclipse, as their platform to build the whole system. They created a visualization tool to see the pictures on Mars. It looks just like Google Earth, but for Mars.
Moreover, the operation engineers in JPL need to evolve lots of custom tools to automate operation processes. Definitely, they have strict processes of designing, implementing, and testing for very high quality software engineering. However, It’s very surprising that they use Perl and Python scripts to do a billion-dollar critical mission, which subverts the general thought that scripting languages are only used for sloppy work.
Upon reflecting on user studies, as prompted by a discussion at UIST, these are three main user study strategies. When are these tests most valuable?
USER STUDY STRATEGIES
Formative User Studies - “What should I do?”
This happens before any substantial design work and can be used to identify needs and problems for a population. This can help us, as researchers, to find real world needs instead of only identifying what we think our needs are. If we don’t understand what future users’ needs are, ethnographies and other formative user studies can be valuable, though time consuming. Time may be better spent looking at and drawing upon others’ previous research. Even unrelated work can help point us toward problems users face regardless of specific applications. However, doing studies like this can help to put researchers into contact with a user base who may be perfect for later evaluations.
One issue with this kind of study is that it can narrow your scope to the point that you eliminate the possibility of discovering serendipitous user bases. This benefit has arisen in several projects to come out of the UID group, and leads us to focus more on the following types of studies. We, as computer scientists, may, after all, be better suited to solve problems instead of finding them.
Iterative User Studies - “What should I change?”
This can be useful in testing prototypes. Steve Krug’s Don’t Make Me Think asserts that with three people in a study, especially one on web design, the researcher can ’squint’ and see most major areas of interest - like finding bugs or ‘falling into usability pitfalls.’ However, this may not be enough people; Nielsen and Landauer use a ‘magic number’ of 5. While there may not be an actual magic number, these small numbers can help find big usability bugs. Larger-scope ‘things’ may not be able to be handled by three people, but “if you’re testing all of Photoshop at once you’re doing something wrong anyway.” Constant iteration and testing of prototypes can help you determine if your device or design can even fit into someone’s life.
Summative User Studies - “How did it turn out?”
Testing and evaluating at the end of development as a method of evaluation is incredibly common and a valuable way to reflect on the benefits of a project as a whole.
Opera collected data on over 3 million web pages to find out how they are being coded. Are they standards compliant? Are they verified by the W3C as standards compliant? Do they contain flash? A selection of interesting web-based statistics found here.
One statistic says that there are more webpages containing a “table” tag than there are sites that contain a “tr” or “td” tag. Are there then lots of empty tables out there on the web? Maybe it’s just that some WYSIWYG HTML editors may allow users to create tables, and not delete the table objects when the user deletes all the rows and columns. Go figure this one out: Opera reports that 33.5% of pages contain Flash. That seems high. Can it all be accounted for by YouTube videos embedded in people’s web pages, MySpace music players and ads?
The IEEE Software Magazine will have a special issue on End-User Software Engineering, and the group is thinking of contributing. The Chickenfoot Firefox extension and the Eclipse Plug-in for issuing Keyword Commands in Java are both examples of work from the group that is relevant (more about these projects can be found here). The deadline is in February 4, 2009
What is the IEEE looking for? A conventional view of end-user software engineering involves helping end-users write correct code. Much end-user programming is done in applications such as Microsoft Excel which do not follow software engineering practices, and may contain bugs that traditional debugging practices doesn’t deal with. In contrast, our work focuses on making it easier for end-users to generate code,rather than encouraging end-users to adopt more rigorous software engineering practices. IEEE may be interested in this different tact on end-user software engineering.
The sharing of knowledge and ideas is of fundamental importance to the advancement of technology. With this goal in mind, MIT’s User Interface Design group meets once a day at Tea Time to brainstorm new ideas, review new technologies and ideas, and share their experiences working in the field.
If we hope to herald innovation by sharing ideas with a research group , then there’s a boundless value to sharing ideas and thoughts with the world at large. With this goal in mind, we will post a daily log of the musings and observations we discuss in our tea time meetings, and welcome your thoughts and comments about Human Computer Interaction, User Interface Design, and increasing the value and effectiveness of how we use technology.
-Clayton