keep in touch: channel, expectation and experience
DESCRIPTION
Presented at CHI 2012, Austin, TX. This work got best paper honorable mention award.TRANSCRIPT
KEEP IN TOUCH: CHANNEL, EXPECTATION AND
EXPERIENCERongrong Wang, Virginia Tech
Francis Quek, Virginia Tech
Deborah Tatar, Virginia Tech
James K.S. Teh, National University of Singapore
Adrian D. Cheok, Keio University, Japan
CHI May 7, 2012
“I’m Hugging You, Dad!”
Was Paul Sending a Coded Message?
Goal
To understand whether and how
remote touch in conjunction with speech
can influence the sense of connectedness
& emotional experience
Touch
Touch & Emotion Communication
Valenced Emotions [Wolff 1963, Brossard 1968]
Emotion Amplifier [Knapp 1997]
Distinct Emotions [Hertenstein 2002]
Context Is Key [Jones 1985]
Touch Conveys Affect in An Immediate Non-symbolic Manner.
Remote Touch
Digitally Mediated Touch
Focused on design
Either by design or user decision
Message Passing
Remote Touch Doesn’t Inherit The Immediacy of Real Touch?
Armband
Shape Memory Alloy
Story-Telling Study
Touch
Point
C1: Audio Only
C2: Audio + Touch Touch
Point
Results shows
Significant reduction in sadness General trend toward greater joviality
Remote Touch can influence the affective components in communication
Wang, Rongrong and Quek, Francis, “Talk & Touch: Contextualize Remote Touch for Affective Conveyance”, TEI 2010
Is the specific modality of touch important?
How does the context influence the effect of remote touch?
Speech + Remote Touch Speech + Flashing Light
Type of Channel
Expectation of Communicative Intent
“Communicative” Remote touch is intentionally sent by the story teller to express
her emotion states.
“Measurement” Remote touch is triggered by a measurement of the story
teller’s emotion state.
Study Design: Channel x Experience
Speech +Additional Channel Speech Only
Channel
Expectation
Remote Touch Flashing Light
Base ConditionCommunicative Communicative Touch(CT)
Communicative Light (CL)
Measurement Measurement Touch(MT)
Measurement Light(ML)
Study Procedure
Story Practice Session Pre-PANAS (Positive Affect Negative Affect Schedules) Story Post-PANAS Semi-Structured Interview
“How close do you feel to the storyteller? Please use the likert scale 1(not at all) to 7(extremely) to indicate.”
“Why do you feel connected this way?”
Data Collection & Analysis
PANAS
Semi-Structured Interview
Audio Recorded (about 4 hours)
Transcribed (90 pages)
Open Coding
Upper-Arm Squeeze Device
Result 1: The Sense of Closeness
CT has the highest rating in the sense of closeness which is significantly higher than Base(p< 0.05)
One Way ANOVA
Communicative
Touch
MeasurementTouch
CommunicativeLight
MeasurementLight
Speech Only
Result 2: Interaction Effect Between Channel and Expectation
Two-Way ANOVA on Closeness Variables
Channel x Expectation
Dependent Variable Closeness Rating
Communicative Measurement
0
1
2
3
4
5
6
5
43.625
4.4375
Interaction Plot
Touch
Flashing Light
A statistically significant
interaction effect (p<0.05)
“Why do you feel close this way? ”
1. “The story is nice.”
2. “The way she told the story.” and “The way she carries her emotion in her voice. ”
3. “Own personal experience makes me feel close.”
4. “The squeeze feels like a personal physical contact and helps me visualize the activity. ” (only in CT and MT)
5. “The flashing light makes me feel more about her feelings.” (only in CL and ML)
6. “Not much in common.” and “I don’t like her story.”
Positive Comments (Category 1~5)
# of positive comments in CT is significantly higher than that in the base condition (p<0.05)
2-Way ANOVA: significant interaction effect between channel and expectation (p<0.05)
Result 3: CT leads to an intensified emotion experience
Changes in Negative Emotions
Changes in Sadness
Changes in Joviality
Conclusion
Context is critical: Recipient’s expectation of communicative intent from storyteller
significantly impacts the sense of connectedness
Results support Immediacy of Touch Affect and connectedness are unlikely to be encoded symbolic
statements
Results support Uniqueness of Touch Difference in touch vs light channels (especially in the interaction
effect) suggest that touch is not easily replaced
Acknowledgement
NSF Keio-NUS CUTE Center Center for HCI at Virginia Tech
Participants Steve Harrison