Introduction to Computer Input

Devices and Their Evaluation

Shumin Zhai
IBM Almaden Research Center
First Mouse Patent (Engelbart, 1964)
First Mouse (Douglas Engelbart and
William English, 1964)
"A Research Center for Augmenting Human Intellect,"
Douglas C. Engelbart, and William K. English, Proc.
1968 Fall Joint Computer Conference
Qualitative Analysis
n Touchscreen
• Pros
• Cons
n Stylus / light pen
• Pros
• Cons
Quantitative Performance Evaluation

n What to measure?
• Depending on the task / application
scenario
n Common measures
• Trial completion time
• Error rate
• Learning speed
• Comfort / fatigue
• etc.
Pointing Device Evaluation
n Real task: Interacting with WIMP
interface
n Experimental task: target acquisition
• abstract, elemental, essential

n Performance measures: time, error rate

Task modeling for evaluation

n Bringtask modeling to device

evaluation
• Card, English, Burr, 1978

“Evaluation of mouse, rate controlled isometric

joystick, step keys and text keys for text selection
on a CRT”,

Ergonomics, vol. 21, 601-613

Fitts’law (Paul Fitts, 1954)
n MT = a + b log2( D +1)
W
ID
D

W
1/b - Index of Performance, Throughput, Bandwidth
Fitts’law
n “The information capacity of the human motor
system in controlling the amplitude of
movement”,

Journal of Experimental Psychology,

vol 47, 381-391
Time (sec)

*
*
*
* *
* *
*
* *
* *
*
* *
* *
*
*
*
ID (bits)
log2(A/W+1)
Experimental Design
n Fairness for the given task
n Wide enough ID combinations
• W’s: from character size (10) to icon (30
pixel)
• A’s: from short (60) to cross screen (800)
n Multiple individuals/subjects ABC
AB
n Balancing orders BCA
BA
n Statistical analysis
CAB

n Controlling error (about 5%)

Lab Assignment
n Measure Fitts’law index of performance with
bare hand on paper
n Measure any two devices using Fitts’law with
the Almaden Program
n Compare performance of the two devices
n Compare devices with bare hand
n Discuss the validity/benefits of Fitts’law in
your study.
n Discuss pros and cons of the devices:
suggest improvements or new designs
Beyond Fitts’law
n Hick’s law
n Key stroke model
n Control theoretic modeling
n Limitations to Fitts law: pointing only
Trajectory-based tasks
ÄExample: hierarchical menus
ÄIs there a “law”to Steering?
Thought experiment...

n2 goals passing A

A
ID = log2 ( +1) W
W
n3 goals passing A/2 A/2

A
ID = 2 log2 ( +1)
2W
n N+1 goals passing A/N A/N A/N

A
ID = N log2 ( +1)
NW
n∞ goals passing A
A
ID = ?
W W
“Steering law”
n Steering law (Accot and Zhai 1997)
• “Beyond Fitts’law: Modeling trajectory based HCI tasks”,
Proc of CHI’97

TC = a + b IDC

IDC =
∫
C
dx
W(x)
Results

W
Device comparison in steering tasks
(Accot & Zhai, CHI’99)
Time
Trackball

Touchpad

Trackpoint

Mouse
Stylus

5 10 15 20 25 30
Steering Index of Difficulty
Conferences and Journals
n CHI: ACM Conference on Human Factors in
Computing Systems
n INTERACT: IFIP Conference on Human Computer
Interaction
n UIST: ACM Symposium on User Interface Software
and Technology
n HFES: Human Factors and Ergonomics Annual
Meeting

n ACM Transactions on Computer Human Interaction

(TOCHI)