際際滷

際際滷Share a Scribd company logo

Cognitive Abilities and Email: Impact of Interface and Task - Dissertation presentation 2004.05.06

J
jacekg

This document summarizes Jacek Gwizdka's research on the impact of email interfaces and tasks on cognitive abilities. It outlines a field study and two controlled lab studies that examined how pending tasks in emails are handled, the effects of different email interfaces on performance, and the role of cognitive abilities like working memory. The research found that interface design, task type, and individual cognitive abilities all impacted email usage and performance. It provided insights into differences in how people manage emails and tasks, and contributions to understanding cognition's role in human-computer interaction.

1 of 21
Download to read offline
2004-05-06 Cognitive Abilities and Email: Impact of Interface and Task Jacek Gwizdka Final Oral Examination 2004-05-06 Interactive Media Lab Knowledge Media Design Institute University of Toronto jgwizdka@acm.org www.gwizdka.com www.emailresearch.org
2004-05-06Jacek Gwizdka 2 Outline Motivation & Background Research Questions Field study ≒TaskView controlled study ≒WebTaskMail controlled study Contributions Future Work
2004-05-06Jacek Gwizdka 3 Motivation & Background Information overload in email Diversity of information in email diverse task Email designed for asynchronous conversations Email not designed for: file transfer & management; contact management; maintenance of social image; personal information management; task and to-do management MB
2004-05-06Jacek Gwizdka 4 Issues Research Opportunities Handling messages related to pending tasks problematic Effects of email interfaces on behaviour little known Role of cognitive abilities in email tasks unexplored More evaluation of email interfaces needed MB
2004-05-06Jacek Gwizdka 5 Research Objectives & Questions More efficient processing of task-laden inboxes How are messages related to pending tasks handled in email? What are the effects of user interface on email performance? How is user performance affected by cognitive abilities? RQ Field Study Two Controlled Lab Studies
2004-05-06Jacek Gwizdka 6 Field Study Findings Pending tasks kept in email User actions compensate for missing functionality (e.g. email to self to keep important tasks on top of inbox) Individual differences : read msg msg after task delay Transfer out of email & Delete Transfer out of email & Keep in email Keep Delete 7 users 7+3 users Keep in email 8 users 1+1 users 4 users Message arrives Future info to PIM applications FS n =19 users do task
2004-05-06Jacek Gwizdka 7 Controlled User Study #1 User tasks information finding : Header task (H) & Date task (D) What are effects of TaskView representation of pending tasks on user performance? What are effects of cognitive abilities on user performance? S1 UI-足Text OutlookUI-足Visual TaskView How can handling of future messages be made more efficient?
2004-05-06Jacek Gwizdka 8 Selected Cognitive Abilities WM - working memory: recall a number of distinct elements for reproduction VM - visual memory: remember location & orientation of visual information FC - flexibility of closure: extract information from distractive background S1
2004-05-06Jacek Gwizdka 9 UI * TASK -足> time for 1st & 2nd session 0 10 20 30 40 50 Header Date Task time sec UI Text 1st session UI Visual 1st session UI Text 2nd session UI Visual 2nd session Effect UI * Task on time Header task faster in UI-Text Date task faster in UI-Visual Results: UI*Task S1
2004-05-06Jacek Gwizdka 10 Results Cognitive Abilities S1 Visual Memory & Task 2nd session: Effect of Task * MV2 on time 0 10 20 30 40 50 Header Date Task time sec MV2 low MV2 high 2nd session: Effect of Task * MV1 on time 0 10 20 30 40 50 Header Date Task time sec MV1 low MV1 high `
2004-05-06Jacek Gwizdka 11 Results Cognitive Abilities (contd) S1 Flexibility of Closure & UI * Task 2nd session: Effect of UI * FC on time for Date task 0 10 20 30 40 50 UI-足Text UI-足Visual UI time sec FC low FC high
2004-05-06Jacek Gwizdka 12 Controlled User Study #2 S2 UI-足Visual UI-足Text Redesigned interface WebTaskMail User tasks: Header, Date & Mixed
2004-05-06Jacek Gwizdka 13 Results Subjective Preferences UI-Visual easier to use than UI-Text UI-Visual preferred for: handling to-do's & events overview of pending tasks
2004-05-06Jacek Gwizdka 14 Results UI Learning (1st session) Learning curve for UI-足Text 00:00 00:10 00:20 00:30 00:40 00:50 01:00 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Question # time sec Learning curve for UI-足Visual 00:00 00:10 00:20 00:30 00:40 00:50 01:00 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Question # time sec S2 Learning effect - both UIs Difference low/high working memory in UI-Visual Low WM High WM Low WM High WM
2004-05-06Jacek Gwizdka 15 2nd session: Effect of UI * Task on time 00:00 00:10 00:20 00:30 00:40 00:50 Header Date Mixed Task time sec UI Text UI Visual Results UI*Task Effect UI * Task on time for 2nd session UI-Text = UI-Visual on Header task UI-Visual faster on Date task UI-Text faster on Mixed task S2
2004-05-06Jacek Gwizdka 16 WM: short-term store & attention control Learning Interaction Performance time Results Roles of Working Memory Effect of Working Memory on time 00:00 00:10 00:20 00:30 00:40 00:50 WM low WM high time sec 1st session 2nd session S2 Effect of Working Memory on sorting 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 WM low WM high sort/Q 1st session 2nd session Learning curve for UI-足Visual 00:00 00:10 00:20 00:30 00:40 00:50 01:00 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Question # time sec Low WM High WM
2004-05-06Jacek Gwizdka 17 Results User Clusters Factors differing between clusters Email Handling Clusters Cluster #1 The Cleaners transfer pending tasks out of email Cluster #2 The Keepers keep pending tasks in email Flexibility of Closure low high S2 Two Email Handling Clusters ( Field study) 1) Transfer pending tasks (7 users) 2) Keep pending tasks (16 users) Differences between clusters:
2004-05-06Jacek Gwizdka 18 Contributions - 1 Effects of cognitive factors on email tasks Effects at different interaction stages WM learning & task performance FC & VM task performance Different performance & interaction measures affected WM & FC time, WM & VM sorting, VM scrolling Opposite direction of effects WM & VM on sorting Multiple roles of working memory in interaction CO
2004-05-06Jacek Gwizdka 19 Contributions - 2 Understanding differences in behaviour In email handling : flexibility of closure & email experience In interaction effort : cognitive abilities (CS, WM, VM) & email experience Methodological contributions Developed email reference task and metrics Demonstrated the effects of tasks on performance CO
2004-05-06Jacek Gwizdka 20 Possible Future Work Narrow down role of cognitive abilities eye-tracker & working memory Field studies UI design personalized and adaptive UI FU
2004-05-06Jacek Gwizdka 21 Acknowledgements l My academic advisor - Professor Mark Chignell l PhD Committee members: Professors R. Baecker, C.D. Sadleir & E. Toms l External Examiner: Professor Chris Neuwirth l KMDI l TimeStore: Professor Ron Baecker & Peter Wolf l My colleague - Dr. David Modjeska l Colleagues from Interactive Media Lab l Field Study @ Xerox PARC Dr. Michelle Baldonado, Ken Pier, and others l This research was financially supported, in part, by NSERC, OGS & BUL

More Related Content

Cognitive Abilities and Email: Impact of Interface and Task - Dissertation presentation 2004.05.06

  • 1. 2004-05-06 Cognitive Abilities and Email: Impact of Interface and Task Jacek Gwizdka Final Oral Examination 2004-05-06 Interactive Media Lab Knowledge Media Design Institute University of Toronto jgwizdka@acm.org www.gwizdka.com www.emailresearch.org
  • 2. 2004-05-06Jacek Gwizdka 2 Outline Motivation & Background Research Questions Field study ≒TaskView controlled study ≒WebTaskMail controlled study Contributions Future Work
  • 3. 2004-05-06Jacek Gwizdka 3 Motivation & Background Information overload in email Diversity of information in email diverse task Email designed for asynchronous conversations Email not designed for: file transfer & management; contact management; maintenance of social image; personal information management; task and to-do management MB
  • 4. 2004-05-06Jacek Gwizdka 4 Issues Research Opportunities Handling messages related to pending tasks problematic Effects of email interfaces on behaviour little known Role of cognitive abilities in email tasks unexplored More evaluation of email interfaces needed MB
  • 5. 2004-05-06Jacek Gwizdka 5 Research Objectives & Questions More efficient processing of task-laden inboxes How are messages related to pending tasks handled in email? What are the effects of user interface on email performance? How is user performance affected by cognitive abilities? RQ Field Study Two Controlled Lab Studies
  • 6. 2004-05-06Jacek Gwizdka 6 Field Study Findings Pending tasks kept in email User actions compensate for missing functionality (e.g. email to self to keep important tasks on top of inbox) Individual differences : read msg msg after task delay Transfer out of email & Delete Transfer out of email & Keep in email Keep Delete 7 users 7+3 users Keep in email 8 users 1+1 users 4 users Message arrives Future info to PIM applications FS n =19 users do task
  • 7. 2004-05-06Jacek Gwizdka 7 Controlled User Study #1 User tasks information finding : Header task (H) & Date task (D) What are effects of TaskView representation of pending tasks on user performance? What are effects of cognitive abilities on user performance? S1 UI-足Text OutlookUI-足Visual TaskView How can handling of future messages be made more efficient?
  • 8. 2004-05-06Jacek Gwizdka 8 Selected Cognitive Abilities WM - working memory: recall a number of distinct elements for reproduction VM - visual memory: remember location & orientation of visual information FC - flexibility of closure: extract information from distractive background S1
  • 9. 2004-05-06Jacek Gwizdka 9 UI * TASK -足> time for 1st & 2nd session 0 10 20 30 40 50 Header Date Task time sec UI Text 1st session UI Visual 1st session UI Text 2nd session UI Visual 2nd session Effect UI * Task on time Header task faster in UI-Text Date task faster in UI-Visual Results: UI*Task S1
  • 10. 2004-05-06Jacek Gwizdka 10 Results Cognitive Abilities S1 Visual Memory & Task 2nd session: Effect of Task * MV2 on time 0 10 20 30 40 50 Header Date Task time sec MV2 low MV2 high 2nd session: Effect of Task * MV1 on time 0 10 20 30 40 50 Header Date Task time sec MV1 low MV1 high `
  • 11. 2004-05-06Jacek Gwizdka 11 Results Cognitive Abilities (contd) S1 Flexibility of Closure & UI * Task 2nd session: Effect of UI * FC on time for Date task 0 10 20 30 40 50 UI-足Text UI-足Visual UI time sec FC low FC high
  • 12. 2004-05-06Jacek Gwizdka 12 Controlled User Study #2 S2 UI-足Visual UI-足Text Redesigned interface WebTaskMail User tasks: Header, Date & Mixed
  • 13. 2004-05-06Jacek Gwizdka 13 Results Subjective Preferences UI-Visual easier to use than UI-Text UI-Visual preferred for: handling to-do's & events overview of pending tasks
  • 14. 2004-05-06Jacek Gwizdka 14 Results UI Learning (1st session) Learning curve for UI-足Text 00:00 00:10 00:20 00:30 00:40 00:50 01:00 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Question # time sec Learning curve for UI-足Visual 00:00 00:10 00:20 00:30 00:40 00:50 01:00 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Question # time sec S2 Learning effect - both UIs Difference low/high working memory in UI-Visual Low WM High WM Low WM High WM
  • 15. 2004-05-06Jacek Gwizdka 15 2nd session: Effect of UI * Task on time 00:00 00:10 00:20 00:30 00:40 00:50 Header Date Mixed Task time sec UI Text UI Visual Results UI*Task Effect UI * Task on time for 2nd session UI-Text = UI-Visual on Header task UI-Visual faster on Date task UI-Text faster on Mixed task S2
  • 16. 2004-05-06Jacek Gwizdka 16 WM: short-term store & attention control Learning Interaction Performance time Results Roles of Working Memory Effect of Working Memory on time 00:00 00:10 00:20 00:30 00:40 00:50 WM low WM high time sec 1st session 2nd session S2 Effect of Working Memory on sorting 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 WM low WM high sort/Q 1st session 2nd session Learning curve for UI-足Visual 00:00 00:10 00:20 00:30 00:40 00:50 01:00 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Question # time sec Low WM High WM
  • 17. 2004-05-06Jacek Gwizdka 17 Results User Clusters Factors differing between clusters Email Handling Clusters Cluster #1 The Cleaners transfer pending tasks out of email Cluster #2 The Keepers keep pending tasks in email Flexibility of Closure low high S2 Two Email Handling Clusters ( Field study) 1) Transfer pending tasks (7 users) 2) Keep pending tasks (16 users) Differences between clusters:
  • 18. 2004-05-06Jacek Gwizdka 18 Contributions - 1 Effects of cognitive factors on email tasks Effects at different interaction stages WM learning & task performance FC & VM task performance Different performance & interaction measures affected WM & FC time, WM & VM sorting, VM scrolling Opposite direction of effects WM & VM on sorting Multiple roles of working memory in interaction CO
  • 19. 2004-05-06Jacek Gwizdka 19 Contributions - 2 Understanding differences in behaviour In email handling : flexibility of closure & email experience In interaction effort : cognitive abilities (CS, WM, VM) & email experience Methodological contributions Developed email reference task and metrics Demonstrated the effects of tasks on performance CO
  • 20. 2004-05-06Jacek Gwizdka 20 Possible Future Work Narrow down role of cognitive abilities eye-tracker & working memory Field studies UI design personalized and adaptive UI FU
  • 21. 2004-05-06Jacek Gwizdka 21 Acknowledgements l My academic advisor - Professor Mark Chignell l PhD Committee members: Professors R. Baecker, C.D. Sadleir & E. Toms l External Examiner: Professor Chris Neuwirth l KMDI l TimeStore: Professor Ron Baecker & Peter Wolf l My colleague - Dr. David Modjeska l Colleagues from Interactive Media Lab l Field Study @ Xerox PARC Dr. Michelle Baldonado, Ken Pier, and others l This research was financially supported, in part, by NSERC, OGS & BUL