Wearable ESM Differences in Experience Sampling Across Wearable - - PowerPoint PPT Presentation

▶
wearable esm differences in experience sampling across
SMART_READER_LITE
LIVE PREVIEW

Wearable ESM Differences in Experience Sampling Across Wearable - - PowerPoint PPT Presentation

Oct 31, 2023 410 likes •784 views

Wearable ESM Differences in Experience Sampling Across Wearable Devices Javier Hernandez (javierhr@mit.edu) * Daniel J. McDuff Christian Infante Pattie Maes Karen Quigley * Rosalind W. Picard Content Background Wearable ESM Tool

slide-1
SLIDE 1

Wearable ESM Differences in Experience Sampling Across Wearable Devices

Javier Hernandez (javierhr@mit.edu) Daniel J. McDuff Christian Infante Pattie Maes Karen Quigley * Rosalind W. Picard

*

slide-2
SLIDE 2

Background Wearable ESM Tool Real-life Experiment Results Conclusions

Content

slide-3
SLIDE 3

Background Wearable ESM Tool Real-life Experiment Results Conclusions

Content

slide-4
SLIDE 4

Experience Sampling Method

(Csikszentmihalyi et al. 1977) (Larson and Csikszentmihalyi, 1983) (EMA, Shiffman et al., 2008; Bolger et al., 2003) Minimize biases Uncontrolled Minimize interactions Natural settings

4

slide-5
SLIDE 5

Social Interactions Time Productivity Emotions Food Intake

Conner et al, 2009 Ayzenberg et al, 2012 Burgin et al, 2012 Scholl et al 2014 Adams et al, 2014 Berkman et al, 2014 Church et al 2014 Timmermann et al, 2015 Hu et al, 2015

ESM: Applications

5

slide-6
SLIDE 6

ESM: Devices

Can we use wearable devices to minimize some

  • f the problems?

Disruptive Missed Interruptions Response Biases

6

slide-7
SLIDE 7

Timmermann et al, 2015

Considered Locations

Stress and mood measurement

Pocket Head-worn

Intimate Very Accessible Cumbersome Very Different Familiar Widespread Disruptive Biases Concealable Accessible Small Display Different

Wrist-worn

Hu et al, 2015 Scholl et al 2014 Adams et al, 2014 Ayzenberg et al, 2012 Berkman et al, 2014 Burgin et al, 2012 Church et al 2014

7

slide-8
SLIDE 8

Background Wearable ESM Tool Real-life Experiment Results Conclusions

Content

slide-9
SLIDE 9

Experience Sampling Tool

2D Grid 5-Likert Scale

9

slide-10
SLIDE 10

VERY PLEASANT VERY UNPLEASANT VERY ENERGETIC LOW ENERGY

How were you feeling during the previous 5 minutes?

slide-11
SLIDE 11

HIGH RESOURCES LOW RESOURCES HIGH DEMANDS LOW DEMANDS

What situation best reflects your previous 5 minutes?

slide-12
SLIDE 12

How stressed were you feeling during the previous 5 minutes?

slide-13
SLIDE 13

Was this prompt disruptive?

slide-14
SLIDE 14

Prompt Triggering

  • 1. Time Distribution. ~1 prompt every 45 minutes
  • 2. Time Variability. Distributed over the day
  • 3. Device Variability. No more than two consecutive prompts on the same device

16

slide-15
SLIDE 15

Background Wearable ESM Tool Real-life Experiment Results Conclusions

Content

slide-16
SLIDE 16

18

slide-17
SLIDE 17

19

slide-18
SLIDE 18

Samsung Galaxy S4 1920x1080 pixels Samsung Gear Live 320x320 pixels Google Glass 640x360 pixels

20

slide-19
SLIDE 19

Starting Phase

  • Tutorial
  • Surveys

Ending Phase

  • Usability of devices
  • Free-form Interview

1st day  $15, 2nd day  $25, 3rd day  $35, 4th day  $45, 5th day  $55 + $25 for successfully completing the whole study $200 for the whole study

Data Collection

21

slide-20
SLIDE 20

Data Overview

15 participants (7 females and 8 males) from MIT Graduate students and one administrator 29.66 years (STD: 6.42) 5 (mostly) consecutive work days

22

slide-21
SLIDE 21

Background Wearable ESM Tool Real-life Experiment Results Conclusions

Content

slide-22
SLIDE 22

Average #Prompts Issued by Device

Response Rate per Person over 5 Days

Total: 627 prompts

* Statistically significant based on ANOVA and Bonferroni correction (p<0.007)

*

24

slide-23
SLIDE 23

Missed: 111 prompts (~18%) Average #Prompts Issued by Device Average (%) Submitted Total: 627 prompts

Response Rate per Person over 5 Days

* Statistically significant based on ANOVA and Bonferroni correction (p<0.007)

*

25

slide-24
SLIDE 24

Response Times

Time (seconds)

Final Submission Prompt Triggered

Total Time

26

slide-25
SLIDE 25

Response Times

Time (seconds)

Final Submission Prompt Triggered 1st interaction with the application

Total Time Starting Time *

27

slide-26
SLIDE 26

*

Response Times

Time (seconds)

Final Submission Prompt Triggered 1st interaction with the application

Total Time Starting Time Answering Time *

28

slide-27
SLIDE 27

Response Distribution

“I found it difficult to point out things [on the Gear] because my finger was on top of it” “The watch was the hardest [to point] because my finger may be too fat”

29

slide-28
SLIDE 28 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 Qua ty o epo ts 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 Very Challenging Very Easy vs 1 2 3 4 5 2 4 6 8 10 Very Inaccurate Very Accurate vs 1 2 3 4 5 2 4 6 8 10 Very Uncomfortable Very Comfortable vs 1 2 3 4 5 2 4 6 8 10 Never All the time vs 1 2 3 4 5 2 4 6 8 10 Not at all Significantly More vs 1 2 3 4 5 2 4 6 8 10 Never All the time vs

Usability of Devices

Difficulty

  • f Use

Device Comfort Elicited Stress Reports Quality Affected Social Interactions Potential Future Use 30

slide-29
SLIDE 29 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 Qua ty o epo ts 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 Very Challenging Very Easy vs 1 2 3 4 5 2 4 6 8 10 Very Inaccurate Very Accurate vs 1 2 3 4 5 2 4 6 8 10 Very Uncomfortable Very Comfortable vs 1 2 3 4 5 2 4 6 8 10 Never All the time vs 1 2 3 4 5 2 4 6 8 10 Not at all Significantly More vs 1 2 3 4 5 2 4 6 8 10 Never All the time vs

Usability of Devices

Difficulty

  • f Use

Device Comfort Elicited Stress Reports Quality Affected Social Interactions Potential Future Use

“[The Glass] is painful, I wear glasses sometimes and they're not that uncomfortable...” “Surprisingly [the Glass] was not uncomfortable”

31

slide-30
SLIDE 30 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 Qua ty o epo ts 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 Very Challenging Very Easy vs 1 2 3 4 5 2 4 6 8 10 Very Inaccurate Very Accurate vs 1 2 3 4 5 2 4 6 8 10 Very Uncomfortable Very Comfortable vs 1 2 3 4 5 2 4 6 8 10 Never All the time vs 1 2 3 4 5 2 4 6 8 10 Not at all Significantly More vs 1 2 3 4 5 2 4 6 8 10 Never All the time vs

Usability of Devices

Difficulty

  • f Use

Device Comfort Elicited Stress Reports Quality Affected Social Interactions Potential Future Use

“People would feel I was taking pictures of them and did not enjoy the conversation” “[The Glass] was a nice ice breaker”

32

slide-31
SLIDE 31 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 Qua ty o epo ts 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 1 2 3 4 5 2 4 6 8 10 Very Challenging Very Easy vs 1 2 3 4 5 2 4 6 8 10 Very Inaccurate Very Accurate vs 1 2 3 4 5 2 4 6 8 10 Very Uncomfortable Very Comfortable vs 1 2 3 4 5 2 4 6 8 10 Never All the time vs 1 2 3 4 5 2 4 6 8 10 Not at all Significantly More vs 1 2 3 4 5 2 4 6 8 10 Never All the time vs

Usability of Devices

Difficulty

  • f Use

Device Comfort Elicited Stress Reports Quality Affected Social Interactions Potential Future Use

“I found it very annoying receiving notification through Glass when speaking with people because it was so noticeable”

33

slide-32
SLIDE 32

Background Wearable ESM Tool Real-life Experiment Results Conclusions

Content

slide-33
SLIDE 33

Overview

Response Time Response Distribution Response Rates Usability

35

slide-34
SLIDE 34
  • Application improvements

Modify submission Self-report on demand

  • Influencing factors

Negative press of Glass Usability of devices

  • Other interactions

Modalities (e.g., voice, camera) Input (e.g., text, numbers)

Limitations & Future Work

36

slide-35
SLIDE 35

Wearable Experience Sampling Tool Qualitative and Quantitative Analysis

Summary

37

slide-36
SLIDE 36

Wearable ESM Differences in Experience Sampling Across Wearable Devices

Javier Hernandez (javierhr@mit.edu) Daniel J. McDuff Christian Infante Pattie Maes Karen Quigley * Rosalind W. Picard

*