UserCentered Interface Design What is usercentered design First

User-Centered Interface Design

What is user-centered design? • First, let’s see some everyday examples of design that is NOT user-centered 2

Classic Stove Example 3

Classic Stove Example Do the controls work this way? 4

Classical Stove Example Or this way? 5

Infamous “butterfly” ballot To vote for the democrats do I punch the second hole or the third hole? 6

“Helpful” color coding Are you sure you want to delete all records from the data base? Yes “Yes” in instance is NOT a good thing No www. iarchitect. com/mshame. htm 7

Many more interesting examples http: //www. iarchitect. com/mshame. htm http: //www. baddesigns. com 8

What is the problem with these designs? • No designer purposes sets out to confuse or irritate(刺激) users • Fallacy(谬论): the most usable design is OBVIOUS by thinking through interior thinking about the problem 9

Reference Information for Java What frame controls what window? 10

But… the target audience loved this approach • The organization was highly meaningful and powerful – It reflects how users think about the information Packages Classes 11

The lesson • Relying solely on interior thinking is a risky approach to design • It risks making incorrect assumptions about the people who will use a system ! 12

User-Centered Design A Focuses on who will use a system from A. the very beginning B Checks in with users each step along the B. way User Research Analysis Design Iterative Design Prototyping & Testing Development 13

User Research Focusing on users from the beginning

User Research Analysis Design Iterative Design Prototyping & Testing Development 15

Discovering User Needs Through Direct Research Observational Studies User Needs User Interviews User Surveys 16

Observational Studies • Observing how people currently perform work within their “live” context Observational Studies User Needs User Interviews User Surveys 17

Observational Studies • People aren’t able to fully articulate（清晰明 白地说） what they do – Behavior is often automatic – Important factors in the social and physical context are taken for granted 18

Example: Observational Study • Kitchen entertainment/information center • Physical context of use is very important – Frequent interruptions – Work occurs in a variety of locations – Cleanliness highly important 19

Example findings (user needs) Context of Use ü User interface for system must be visible and controllable from a distance ü Touch screen has limitations for use – Hands are frequently wet and/or dirty – Stylus（铁笔） must be attached to unit 20

User Interviews • Asking users to describe how they perform work and what they think about it Observational Studies User Needs User Interviews User Surveys 21

User Interviews • Less open-ended（自由回答） than observational studies – Questions usually focus on getting information about particular user goals and tasks • Best conducted in the environment where users perform work 22

Example: User Interviews • Web Site targeted at software developers – How can we improve the information we provide to developers? – Questions about how they seek information 23

Example findings (user needs) ü Concentrate information on fewer pages – To leverage using browser search function to find information ü Provide fresh technology news daily – Very consistent pattern of proactive information seeking 24

User Surveys • Asking people specific questions on use and satisfaction Observational Studies User Needs User Interviews User Surveys 25

User Surveys • The Internet makes these easy and inexpensive to deploy • Good for getting specific data on existing features 26

Example: User Survey • Developer Web site – Used to get more data on specific issues identified in open-ended interviews 27

Example findings (user needs) ü Optimize PDFs for on-screen use – PDF Printouts used less frequently ü Replacing search engine is not a high priority – Satisfaction level with current search technology is relatively high 28

User Research: Potential Pitfalls（缺陷） • Methods misapplied or executed poorly • Results may point the wrong direction 29

Potential Pitfall: Just Asking People What They Want • Here’s what Homer Simpson says he wants in a car: – “I want a horn here, and here. You can never find a horn when you‘re mad. And they should all play ‘La Cucaracha（一种墨西哥歌舞） ’. ” 30

Pitfall: Asking People What They Want • Think of the doctor analogy – People are expert in understanding what the symptoms are, but not the underlying disease or how to treat it – The user researcher must be skilled at diagnosis 31

Prototyping & Testing Checking in with users each step along the way

Prototyping & Testing User Research Analysis Design Iterative Design Prototyping & Testing Development 33

Early prototypes are low fidelity（逼真度） • They may be very exploratory designs • Mostly, paper prototypes are used – Pages may be hand-drawn or computer-drawn • Goal: Get user feedback early when changes are easy to make 34

Low Fidelity vs. Final Design 35

Usability Testing of Prototypes 1. Create task scenarios 2. Recruit participants to perform the tasks using the prototype 3. Use the results of testing to fix any problems 4. Repeat Steps 2 -3 36

Conducting test sessions • Usually involves the “Think Aloud Protocol” (TAP) – User expresses thoughts out loud as they perform tasks – Clip shows testing a system in the final design stages 37

Usability Testing with Paper Prototypes • Participants “click” by pointing • They enter data by writing on the page Picture-in-picture video captures people’s actions, what they say, and their facial expressions 38

In the TAP the facilitator must remain neutral • Key to the success of the method • Participants will look to the facilitator for reassurance – May feel like they are “failing” the test – Some participants become highly stressed – Sessions can be long and unpredictable 39

How many users do you need to test? • 6 -8 users • Clients are initially skeptical • But after watching the first set of sessions they believe – You start seeing the same problems again & again – Conserve your participant budget for additional rounds of testing 40

How much of the system do you test? • Some testing is always better than none • There always time and budget tradeoffs • Prioritize the most risky areas of a design – Or focus on “mission-critical” areas such as the check-out flow for e-commerce sites 41

“Discount” Usability Methods

Reality Bites… • The time and expertise for user-centered design may be absent on many projects – Small projects – Engineering-driven company culture • Focus on internal thinking to solve design problems • “We know what users want already” 43

Options • Conduct a heuristic evaluation – Heuristics represent user interface design best practices – Jakob Nielsen is a good starting point www. useit. com/papers/heuristic/ 44

Options • Test with even one or two users – Nielsen: zero users yields zero insights http: //www. useit. com/alertbox/20000319. html 45

Options • Test with internal people – They should have no involvement with the design – Get folks close to the target audience as possible • Risk of misleading information 46

Learning more about User. Centered Design and Usability

Some Books • The Elements of User Experience: User. Centered Design for the Web – James Jesse Garrett • User-Centered Design: An Integrated Approach – Karel Vredenburg, et al 48

A Few Web Resources • Jakob Nielsen – www. useit. com/ • Usable Web – www. usableweb. com/ • IBM - Ease of Use – www-3. ibm. com/ibm/easy/eou_ext. nsf/Publish/570 49