CSC 8570 USI

1
CSC 8570 -- USI

• Class Meeting 3
• September 12, 2007

2
Homework for Today

• Research team membership
• Finished
• Research areas
• Comments sent
• GOMS analysis
• Hand in

3
Creating Tables

• Concerns
• 7 of 14 tables didnt follow specifications and
  example row
• Several examples of header row (not required by
  specs) poorly formatted or with less helpful
  labels

4
Interacting with the Interface

• Issues are
• Types of interface widgets
• Methods of interaction
• Formal notation for interactions
• Extensions to the Keystroke Level Model
• Timings for each simple interaction
• Interactions for control vs. for content
• Quasimodes
• Others?

5
Interface Examples

• Focus on battery operated devices on our first
  weeks list and other small interfaces
• Remote car lock control
• Watch
• Cell phone
• IPod
• Laptop computer
• PDA
• Programmable thermostat
• Simple calculator

6
Interface Examples (2)
7
Notation

• How should we describe interactions with (or
  actions taken with) the interface devices?
• Time for some discovery groups

8
Group Work

• Pair up by threes, except one pair of two.
• Create a list of simple (atomic?) widgets that
  appear in interfaces. Consult the list of
  examples, generalize, then synthesize.
• For each item on the list, state the actions that
  can be performed on it.
• Suggest a notation for each action expressed in a
  way that it is easy to form a sequence of actions.

9
Group Work (2)

• Notes
• Distinguish between physical devices and actions
  on them and interface widgets and actions on
  them. Both types should be on your list.

10
Group Work Summary

• Construct the union table of widgets and actions
  on them.
• Invent the formal notation for describing the
  actions.
• Create some examples of performing tasks using
  these widgets, expressing the method in the
  formal notation.

11
Interface Widgets

• Physical devices
• Keyboard
• Graphical Interface Device (GID)
• Point, buttons
• Touch pad
• Gestures, point, tap
• Virtual devices
• Programmed as objects and methods
• Summary

12
Interface Widgets (2)

• The linked chart shows the summary of the group
  work.
• The chart will be revised as
• new widgets are created or discovered
• new actions are needed
• new notation is invented

13
Interface Models

• Look at
• Physical interface
• Operations of physical interface
• Task set
• Error cost

14
Error Cost

• Definition The cost, measured in time or
  keystroke level actions, needed to recover from
  an erroneous action with a widget.
• Example Mistyping the as teh. Error cost is
  four keystrokes 2 backspace and 2 for the
  correct keys.

15
Models of Physical Action

• Fitts Law moving a cursor to a target
• Keystroke Level (Card Moran) timing
  predictions for a task represented as a sequence
  of physical operations
• Hicks Law time for choosing among a number of
  widgets

16
Fitts Law Questions

• How does Fitts Law generalize to 2-dimensional
  targets?
• Where does the user aim?
• Where does the user land?
• What path is followed?
• What happens with partially hidden targets?

17
Notations for Actions

• Actions for a keyboard and graphical input device
  (GID) include
• Tap pressing and immediately releasing a key.
  Implemented with a momentary contact
  (spring-loaded) switch.
• Click Position the GID and tap the GID button.
• Double click Position the GID and tap the GID
  button twice quickly, without repositioning.

18
Notation for Actions (2)

• Keystroke Level Model
• K , a keystroke. E.g. Kb means tapping lower
  case b
• P, pointing. E.g. positioning the GID
• H, homing. E.g. moving from keyboard to GID or
  GID to keyboard
• M, mental processing
• R, responding

19
Notations for Action (3)

• What does the description
• To create a right justified paragraph, use
  Control .
• ask you to do?

20
Notations for Action (4)

• More actions
• Press and hold a key
• Release a key
• But on what signal or input?

21
Notations and Actions (5)

• Raskin/Beck
• Press and hold bi
• Release bh
• Tap bibh, shortened to bih or just b
• Release on signal bh
• Note that key names are italicized
• Examples
• Create an upper case M ShiftimimhShifth,
  shortened to Shiftimihh

22
Notations for Action (6)

• Examples (continued)
• Distinguish between Enter and E n t e r Enter is
  the name of a key and E n t e r is short for
  Eihnihtiheihrih
• bih20 gives a string of bs
• CtrliAltiDelihhh brings up the Task Manager

23
GOMS

• Example
• Goal Selecting an item from a menu
• Method
• Find menu heading
• Point cursor at menu heading
• Click mouse button (displaying menu entries)
• Find desired entry
• Point cursor at menu entry
• Click mouse button (choosing menu option)

24
GOMS (2)

• Assume that you manage your email messages using
  a hierarchy of folders that is at least five
  levels deep. Develop a GOMS analysis for
  deleting a message in a level three folder when
  you currently are reading messages in your inbox
  (a level one folder).

25
GOMS (3)

• Goal
• Operators
• Methods
• Selection rule

26
Research Project

• Teams (done)
• Areas (should be narrowed, can be revised)
• Hypothesis
• Independent variables
• Dependent variables

27
Research Project (2)

• Sources
• ACM Digital Library
• Science Citation Index
• Web, but be careful
• Conference proceedings, MSC 159
• Bibliography
• Entered into EndNote

28
Research Project (3)

• IRB Form
• Consent form

29
Research Project (4)

• Experimental design
• Between subjects
• Within subjects
• Sources of bias
• Sources of subjects
• Logic of conclusions

30
Next Time

• Research project activities
• Create a GOMS model for constructing the table of
  sums of powers of 2. The method must be at the
  keystroke level and include the 45 keystrokes
  necessary to enter the data.

31
Next Time (2)

• Research project
• Submit hypothesis, independent variables,
  dependent variables
• Submit draft of IRB form
• Submit bibliography as printed from EndNote

32
The Future of HCI

• John Canny, The Future of HCI, from HCI, 4(2006),
  July/August
• Do Cannys views match our ideas?
• What did he miss?
• What did we miss?

33
Research Team Meetings