CSC 8570 USI

1
CSC 8570 -- USI

• Class Meeting 4
• September 19, 2007

2
Topics of the Day

• GOMS review
• Bunt, et al. Understanding the questions raised
  and the answers given.
• Good and bad interfaces
• Research project issues

3
Hand In

• Everyone GOMS model of table creation
• Each research team
• Hypotheses, independent and dependent variables
• First draft of IRB form
• Printout from EndNote of research bibliography

4
GOMS

• Goal
• Operators
• Methods each method is a sequence of operators
• Selection Rule

5
GOMS Examples

• Deleting a message from a mail system folder
• Unlocking a car with a remote control

6
Clock Setting

• Set the clock one hour earlier, the common action
  at the end of daylight savings time.
• GOMS model
• Goal stated above
• Operations
• Methods
• Selection Rule

7
Bunt, Conati, McGrenere

• Pair off by twos
• Create a list of concepts that need further
  explanation, phrases that need clearer
  definitions, and descriptions that are just plain
  mysterious.
• Interpret the assigned experiment

8
Bunt (2)

• Adaptable Users can customize the interface to
  meet their needs
• But users rarely move from default settings
• Adaptive Model users interests, preferences,
  and use characteristics then interface changes
  to support this model.
• But users suffer lack of control, transparency,
  and predictability

9
Bunt (3)

• Adaptive issues
• Is customization worth the effort?
• Can users customize effectively?
• Does the system provide support (and help) for
  customization?

10
Bunt (4)

• Experiment 1
• Hypotheses?
• Independent variables?
• Dependent variables?
• Subjects?
• Statistical analysis?

11
Bunt (5)

• Experiment 2
• Hypotheses?
• Independent variables?
• Dependent variables?
• Subjects?
• Statistical analysis?

12
Bunt (6)

• Suggest some experiments that follow from this
  work.

13
Good and Bad Interfaces

• Who says?
• On what basis?

14
Good and Bad (2)

• What do users want?
• How do designers know?
• Are there fundamental design principles that
  should be followed?
• For example?
• Do designs match mental models?
• What if they dont?

15
Good and Bad (3)

• Generative User Engineering Principles (GUEPs)
• Cognitive Dimensions (CDs)
• of notation
• of representation in an interface
• of understanding

16
Research Project

• Hypotheses (tonight)
• Independent variables (tonight)
• Dependent variables (tonight)
• Bibliography
• First entries (tonight)
• Continuous process
• Background
• Formal setting (lets talk)

17
Theory Collection

• Visual processing
• Three-stage visual system
• Preattentive processing theory
• Structured object perception theory (Biederman)
• Scientific color theory (CIE)

18
Theory Collection (2)

• Motor Behavior Models
• Hick-Hyman Law choice time
• Keystroke-Level Model error-free task completion
  time
• Three-state model of graphical input (Buxton)
• Fitts Law human movement
• Guiards Model of Bimanual Skill

19
Theory Collection (3)

• Task analysis
• Hierarchical model using ConcurTaskTrees
• Design principles
• GUEPs
• Cognitive Dimensions
• Mental models

20
Research Project (2)

• Short presentation
• One spokesperson
• Five minutes
• Hypothesis, variables, initial experimental
  design
• PowerPoint

21
Next Time

• Read Edge and Blackwell, Cognitive Dimensions
  Tradeoffs in Tangible User Interface Design
• You find this paper by going to
  www.cl.cam.ac.uk/afb21/CognitiveDimensions
• Then choosing the link for the 10th Anniversary
  Workshop on CDs research
• Create a concept map relating the ideas of the
  paper to one another
• Use the concept map software
• Make sure you include at least 12 concepts and
  maybe many more
• Consult the papers by Thomas Green for more
  information
• Research project presentation

22
Research Team Meetings