Upload
hugo-nicolau
View
452
Download
3
Embed Size (px)
DESCRIPTION
Presentation @ CHI'12
Citation preview
Touch Typing using Thumbs: Understanding the Effect ofMobility and Hand Posture
Hugo [email protected]
Joaquim Jorge
SIIDSituationally-Induced Impairments and Disabilities
Hand oscillation hinder performance[Bergstrom-Lehtovirta et al., 2011]
Especially with small virtual targets[Mizobuchi et al., 2005; Lin et al., 2007]
Thumb interaction
Text-entry whilst walking
Effects of walking and hand posture on touch typing
22 participants19 males, 3 females
23 to 40 years old
Right-handed
Only 15used mobile touchscreens
regularly
3 mobility conditions
Seated Walking 65% of human pace[Barnard et al., 2005]
WalkingAverage human pace2 steps / second
Indoor test track and Pacesetter
Fixed pace [Kane et al., 2008]
Comparable level of walking demand
Vibrotactile feedback to pacesetter
Hand Postures
One-handportrait
Two-handlandscape
Two-handportrait
Copy task
2 practice sentences / condition
5 test sentences / condition
Portuguese language representative corpus
Error correction was not available
Apparatus
HTC Desire
7x10 mm – portrait mode
10x10 mm – landscape mode
Neither word prediction nor correction was used
Experiment Design
22 participants x 3 mobility settings x 3 hand postures
990 sentences overall
Two-way repeated measures analysis
Results
Two-hand input is faster
19.9
23.928.1
20.5
25.029.1
20.0
24.428.9
0
5
10
15
20
25
30
35
One-hand portrait Two-hand portrait Two-hand landscape
Words per Minute
Seated Walking (65%) Walking (100%)
Two-hand posture does not provide additional stability
0%
2%
4%
6%
8%
10%
12%
14%
16%
0.0 g 0.5 g 1.0 g 1.5 g 2.0 g
MSD Error Rate by Hand Oscillation
One-Hand Portrait Two-Hand Portrait Two-Hand Landscape
seated walking (65%) walking (100%)
Substitutions are the most common type of error
4.3% 3.8%1.7%
1.1%1.1%
0.7%
0.6% 1.1%
1.0%4.8% 5.2%
1.8%
0.9% 1.2%
0.8%
0.8%0.9%
0.8%7.0% 5.5%3.0%
1.0%1.4%
1.0%
0.6% 1.8%
1.2%
0%
2%
4%
6%
8%
Del Ins Subs
Hand oscillation causes poor aiming nor finger slips
88% 84%87%92% 87% 93%93%
81%
96%
0%
20%
40%
60%
80%
100%
120%
One-hand portrait Two-hand portrait Two-hand landscape
Substitutions: incorrect land-on
Seated Walking (65%) Walking (100%)
Substitution pattern
Same-row errors
Distance of one key
Typically at the right
Pattern was consistent across mobility demands
Most users prefer larger keys
0
2
4
6
8
10
12
14
16
18
One-hand portrait Two-hand portrait Two-hand landscape
Participants Preference
Preferred hand posture Less preferred hand posture
Design Implications(For all those sleeping/bored … WAKE UP!)
!Do not over rely on two-hand interaction for physical stability
~
!Take advantage of typing behaviors
T Ytranscribed > payyernintended > pattern
!Design for poor aiming, especially whilst mobile
Touch
Accelerometer data
+ Intended key
Conclusion
Negative effect of walking on touchtyping using thumbs
Consistent substitution pattern
Two-hand interaction does not improve text-entry accuracy; Increasing target size is an effective solution
Design implications
The End.
Hugo [email protected]://web.ist.utl.pt/hugo.nicolau/