This site contains archived material that may be obsolete. For more information about AIGA please visit us at www.aiga.org.
 

LOOP: AIGA Journal of Interaction Design Education
June 2003 Number 7

Results
 

Introduction

Jumping in the <nobr>Deep End</nobr>

User Research

Task and Information Models

Four of the class’s final presentations are available for download. Here are some highlights.

Ross Carl’s CarHounder: one interface won’t do
Ross Carl looks for ways to help people search for cars on the web. His design is based on a key research insight: while there are a lot of attributes one might use to fully describe a car, different people care about different subsets of those attributes. People who want a really cheap ride put price first and might actually be attracted to a little bit of wear and tear so long as the essentials (like good tires) are in place. But someone who is buying for prestige or speed will put other attributes, like horsepower or leather upholstery, ahead of price.
      Ross’ interface begins by asking people to identify with a “qualitative category” then adjusts the information, imagery and controls accordingly. His prototype doesn’t fully explore the opportunities for visual variation (color, type, layout, etc.), but it serves to test the concept of this kind of variation. Ross also explores different controls: drop-down menus with choices of features and stats for technology-oriented speed junkies versus “about this much” sliders for people looking for a car that reflects their personality.

Burstein, Kelkar and Seop: not searching, but collecting

 

 

- Ross Carl 1

-

Ross Carl 2
Variation in Ross Carl’s design. Specifying search parameters for a fast car (top) and a personality car (bottom). See more by downloading Ross’ final assignment paper.

carl.pdf (7.5M PDF document)

 
-
Burstein et al 1

-

Burstein et al 2
Burstein, Kelkar and Seop’s Albums. Top: Examining a list of Chicago neighborhoods after putting one in our album. Bottom: With three neighborhoods in our album, we begin collecting apartments. See more by downloading the final assignment paper.

burstein-kelkar-seop.pdf (1.8M PDF document)

 

The team of Adrian Burstein, Anjali Kelkar and Cho Seop looks at how people search for apartments. They discover behavior less like searching and more like hunting-and-gathering: buyers wander through possibilities within desirable parameters, collect candidates then plan outings to visit several apartments at a time. This process flips back and forth between neighborhoods and apartments, as findings in one trigger inquiries into the other.
      Their solution lets people explore the qualities of neighborhoods that interest them (location, schools, crime rate, etc.) and gather ones they like into an “album.” Within each neighborhood, they can explore the available apartments and add ones that interest them to the album. While most search interfaces simply present a list of search results, this design considers what people do with the results. By presenting “hits” as parts that can be added to an album, the interface lets people build the organization they need when looking at apartments: a detailed list of promising candidates organized by location.

Kim and Song: sequential navigation filters information

Eunjoo Kim and Ki-Bok Song’s design reflects thoughtful consideration about how to deal with complexity. There are so many apartments, each with a large number of attributes. Their solution uses a four-step sequence as primary navigation. Relevant slices of information became visible at each step. Each step has an associated primary view with view-specific and search controls on the left. That is, since people have different criteria at each step of the process, the views flexibly present different data. The flexibility is made explicit in the controls.

Robert Zolna: visual comparison
Robert’s project also compares apartments. His research and design focuses mainly on what people look for and think about when they compare candidates. He provides information visualizations about attributes of candidate apartments and uses multiple views to support different questions or priorities. With Robert’s design, people use their visual sense—evaluating color, position, size and pattern—more than they read. This is an unusual quality for search interfaces. There is a woeful lack of attention to presenting search results for use as compared to the dense text listings and shadowy photographs available on most commercial websites.

 

- Kim and Song 1

-

Kim and Song 2
Kim and Song: mapping info to task. Top: In the first step, "search," and we play with a distance-circle to reveal apartments within a particular location. Bottom: In the "compare" step, our possible apartments are presented in parallel; controls on the left show only the user’s desired attributes for the sake of comparison. See more by downloading the final assignment paper.

kim+song.pdf (1.4M PDF document)

 
- Robert Zolna’s visualizations  
Robert Zolna 1
-
Robert Zolna 2
- Robert Zolna 3
- Robert Zolna 4

 

My thanks to these students and many others who aren’t named in this article for their hard work, curiosity, cheerful engagement and inspiring dedication to quality.

 

 

Also See
 

Adrian Burstein, Anjali Kelkar and Cho Seop. <www.spacefinders>. burstein-kelkar-seop.pdf (1.8M)

Ross Carl. A Qualitative Search Engine for Automobiles. <www.carhounder.com> carl.pdf (7.5M)

Peter J. Denning, Pamela A. Dargan. "A Discipline of Software Architecture." Interactions 1 (1994).

Eunjoo Kim and Ki-Bok Song. Interface Design for Searching Apartments. kim+song.pdf (1.5M)

Slobodan Kalajdziski. UML in Seven Days.
<odl-skopje.etf.ukim.edu.mk/uml-help/>.

Marc Rettig. "Prototyping for Tiny Fingers." Communications of the ACM. 4 (1994).

Robert Zolna. Comparing Apartments. zolna.pdf (727Kócomments in red describe findings from user tests)

 

LOOP June 2003 Number 7