Usability Beyond the Desktop

1
Usability Beyond the Desktop

• Loren Terveen
• CS 5115, Fall 2007
• November 5

2
Announcements

• No TA meetings next week and deliverables
  cancelled
  
• Next week class will meet in the Usability Lab
  basement of Walter Library
  
• Monday, Tuesday, Wednesday
• 230 400

3
Hall of Fame/Shame

• Today
• Jon Alstrup Selam Rodriguez
• Matt Broten and Chris Curtis
• Wednesday
• Dan Anderson Adam MomsenJohn Navil Joseph
  Robert Hosman

4
Hall of Fame/Shame Presentation
MDA vs. iPhone Jon Alstrup Selam Rodrigue
z
4
5
T-Mobile MDA vs. Apple iPhone
5
5
6
T-Mobile MDA Hall of Shame?

• Poor Visibility
• Buttons are very small and hard to push. Stylus
  is necessary most of the time and difficult to
  use
  
• Many common features are not easily
  visible/accessible
  
• Program accessibility is limited to traditional
  windows mobile start menu
  
• Gulf of Execution/errors
• Button options
• mail and explorer
• Communication Manager
• Voice Dialing

6
6
7

• Mappings
• Left and right physical buttons mapped to screen
  buttons
  
• Directional Pad mapped to directions
• Horizontal layout when sliding physical keyboard
  
  
• Constraints
• When in sleep mode, prevents user from
  accidentally hitting buttons
  
• Standardization
• Qwerty keyboard layout
• Volume control location

7
7
8
Apple iPhone Hall of Fame

• Visibility
• Most common functions are visible from the main
  screen
  
• Large buttons that are easy to press
• Affordances
• Affords touching
• Buttons
• Scrolling
• Sliders
• Other
• Mappings
• Movement of your finger, maps to actions on the
  screen
  

8
8
9

• Standardization
• Qwerty keyboard layout.
• Volume up/down buttons on side of phone.
• Icons
• Constraints
• When in sleep mode, prevents user from
  accidentally hitting buttons.
  
• Few physical buttons.
• Preventing Errors
• Consistent Menus
• Consistent location of the same function
• Size of the buttons
• Precision of the touch screen

9
9
10
Car Part SitesStrike BackBrought to you
byMatt Broten Chris Curtis
11
Interface 1 Ecklers Late Great Chevy
Specify year
Quick search option
Main navigational area
12
Interface 2 Rock Auto
Main navigational area
Quick Part Number search option
Helper search option
13
Gulf of Execution
Very Low conceptual model is similar to navigat
ing a directory tree for specific files the
user is guided through the process of finding a
specific part, then results are presented
Search field is empty and clearly identifiable
purpose is clearly understood (find category)

• Moderate
• Provides conceptual model consistent with other
  shopping sites navigation uses links,
  drop-downs, search, and etc.
  
• - User is distracted with results immediately,
  this leads to wasted time looking for parts
  early
  
• - Search field is filled with text, it takes
  longer for the user to determine it is a text
  field

14
Gulf of Evaluation

• Moderate/Low
• user can determine where they are by looking at
  the linking structure not as easy to determine
  depth

Low user can see where they are in the director
y tree from this they can determine exactly what
the displayed results are for and the exact depth
15
Constraints

• Good
• the search filter options are semantically
  constraining (all parts are contained in a
  specific subgroup)

Good only make, model, years, and etc. that hav
e parts available are displayed to the user.
16
Visibility

• Moderate
• status linking structure allows the user to
  see exactly where they are
  
• - Filtering options are on the left side of the
  screen out of the primary focus of the user,
  due to the results shown in the center of the
  screen

High status tree allows the user to see exactl
y where they are all available options are visi
ble to the user when the folders are expanded
folder icon and plus/minus symbol indicate
whether the folder can be expanded/contracted
17
Feedback

• High
• updated results are displayed as the user
  refines their search

High visual feedback is given for every user ac
tion. e.g. search gives the user instant feedback
. If no search results are found, user is
informed otherwise, categories containing search
results are highlighted.
18
Error Prevention/Reduction(preventing the user
from buying the wrong part)

• Moderate
• product image and descriptions are displayed to
  the user
  
• -user can purchase a part without providing
  specific details

High the user is forced to provide detailed spe
cifications before results are displayed and
purchased product image and descriptions are di
splayed to the user
19
The Comparison

• Rock Auto is the winner!

20
Any Questions?
21
Agenda

• Review usability issues and methods covered so
  far
  
• Begin to step back
• From UI design and usability to design-oriented
  HCI research
  
• Beyond desktop interfaces
• Usability issues and methods
• Examples

22
Review
23
What are you trying to learn (1)?

• Qualitative experiences of usability
• What will they use this thing for anyway?
• Trouble spots in completing tasks
• Features found / not found
• Reactions to design elements/decisions
• Learning users mental models
• Why cant those silly users do it?
• Process!

Early Formative evaluation
Most relevant to this class / your project
24
What are you trying to learn (2)?

• Concrete, quantitative measures of usability
• Time to learn a feature
• Use time for specific tasks and users
• Features used (or not)
• Error rates
• Measures of user satisfaction
• Comparison to prior/alternative versions,
  competitors
  
• Results!

Later Summative evaluation
Research-y, but I will argue for its relevance
25
Another quick example

• Pie menus
• An alternative menu design
• Some intuitive and theoretical basis for thinking
  this design would be better
  
• How to test?

Method Controlled (Lab) Experiment
26
Yet another quick example

• Tagging in MovieLens
• Questions
• What influences users in how they apply tags?
• Their own prior usage
• Community usage
• As mediated by a selection algorithm

Method Field Experiment
27
Generally

• Know your question
• Know your burden
• Know the right method