When the Future Becomes the Past:

1
When the Future Becomes the Past
A Study of PIM and Email Use in Time
(Submission as CHI2000 Short paper by Saturday
afternoon)
Jacek Gwizdka Interactive Media Group, UT(M)IE
2
Presentation Outline

• Motivation
• Related work
• Method
• Results
• Conclusions and future research

3
Motivation - Background

• Information overload ? Lack of information
• Pieces of information are buried across many
  computer applications and other media (e.g.
  paper)
• Locating information often requires using several
  tools
• (tool a computer application, a paper-based
  tools)
• Organizing information is difficult
• Tools are used in unexpected ways
• Email is becoming pervasive

4
Motivation First idea

• PARCs DocuShare project building actions into
  paper documents active paper documents
• Notable annotating objects (docs) using
  handhelds
• Use paper calendar to control distribution of
  meeting documents.
• Link calendar entries to notes taken using
  Notable during meetings
• Does it matter what is being linked and in which
  order?
• Calendar is used as an external prospective
  memory
• Note-taking tools are used as an retrospective
  memory
• Notes are taken during an event with an intention
  of being used after the event.
• Can this observation be generalized? Is this type
  of usage enforced by tools? Or is it traditional?

5
Research Objectives

• Focus on personal information use in time
  defined by the following types of information
• Prospective future
• Ephemeral current short-term information
• Working current medium-term information
• Retrospective - past
• How information is transferred between different
  media (paper, electronic) and applications
• What media and applications are used as
• Information carriers
• Working media
• Archival media

6
Research Objectives my PhD

• PhD ? Study life-cycle and temporal attributes of
  email messages
• How does email fit into the larger picture of
  personal information management tools?
• Is information transferred between email messages
  and other PIM tools? How? When?

7
Related Work

• Personal information organization studies
• Malone (1983), Baecker, et al (1994)
• B. Nardi T. Erickson, Barreu Nardi
• V. Bellotti and I. Smith Raton Laveur (1999)
• Email studies
• Mackay (1989), Whittaker, A. Lantz, O. Bälter
• Research Systems employing time as an
  organizational principle
• Lifestreams (Gelernter, Freeman) sending
  reminders to the future
• TimeStore (R. Baecker, Jovicic, Yiu, Silver,
  Long)
• Oberon and Puck (PARC B. Walker 1999)

8
Related Work - continued

• Others studied
• how people organize information in their work
  environments and on their desktops email use and
  email overload time as an organizational
  principle how to integrate event-based
  information with documents.
• Past studies did not focus on the use of PIM
  tools in time
• Lack of studies of relations and links between
  different tools and of information transition in
  time between the tools

9
Study Method

• Small sample redesign and iterate
• 4 pilot subjects and 8 real subjects
• Questionnaire (25 Qs) and a personality
  questionnaire on the web (Keirsey based on
  Myers-Briggs typology)
• In-person interview (30-60 minutes) in subjects
  office (with two exceptions), taped
  (not-transcribed yet)
• Subjects - different roles
• administrative secretary, 3 business owners
  (finance, web site design, hi-tech), 2 grad
  students (engineer-manager, interface designer),
  2 consultants-advisors (research info. viz.)

10
Analysis of Results

• Huge diversity of organizational environments,
  roles and personal styles a list of possible
  uses, rather than frequency of use analysis
• Perspective of a tool provider satisfying needs
  of a diverse customer base.

11
Results

• Information transitions and use of tools as
• Information carriers in- and outbound
• Working media
• Archival media
• (media paper tools or electronic tools and
  applications)
• Linking related information
• (link related, belongs together)
• Email handling and prospective use

12
Results info. transitions

• Manual transitions require either a separate data
  entry or copypaste
• PDA syncing process is an example of
  automatically handled transitions between media
  (and apps?)
• PDA syncing tools make manual transitions easy
  dragdrop icons, but icons (Word, Excel,
  clipboard) are available for addresses, to-do,
  memos, and NOT for calendar

13
Results use of tools

14
Results Linking information

• Gross categories projects, clients, courses
  folders based on categories kind of explicit
  links
• Links between documents and events
• No explicit links even when possible
• Implicit links created by contextual information
  recorded and remembered

15
Results Email handling

• 3 main choices delete handle now handle
  later
• Handle than file/delete - Messages stay in Inbox
  until they are all handled, afterwards they are
• filed in email folders
• some filed in file system folders
• stay in Inbox as read
• deleted
• File than handle - Messages are manually or
  automatically (filters) filed into email folders
  where they are handled (in addition to Inbox)
• this approach gives structure to handling email
  (struct to-do), folders help to prioritize email
  handling by project/role
• used only by one subject, surprising?

16
Results Email handling cont.

• File than handle is rare, surprising? - perhaps
  not
• creating and maintaining email filters is
  demanding
• difficult to get a visual overview of to-do -
  messages distributed among several folders
• Handle later
• To-do (real to-do, or email to-do) ? Email
  use as a prospective tool

17
Results prospective use

• Timed information
• Un-timed information (general to-do)
• Desired features overview, reminders, visible
• Email use as a prospective tool imposes a lot of
  cognitive overhead
• manually going through lists of email messages
• distinguishing timed and un-timed messages
• Strategies flagging messages
• Timed flags?
• Ephemeral information dragging out to desktop

18
Conclusions

• What integration?
• Use of a given tools is often decided based on a
  user-preferred feature (an overview, a
  hierarchical organization of to-dos, very
  idiosyncratic. Tools should offer great
  flexibility.
• Traditional use of tools very strong e.g.
  calendar
• No strong evidence that people use events to
  organize other documents around them. Event
  information used mostly in information retrieval.

19
Conclusions continued

• PIM tools used according to traditional divisions
  of functionality
• calendars to-dos prospective
• note-taking tools retrospective
• Email is a true chameleon it is used, across
  subjects, in all time-roles and both as
  information carrier, as a working and archival
  media
• Often for purposes it was never designed for
  (e.g. ephemeral information)

20
Future Research

• There will be future research
• Long term study of email use

21
PhD Research Objectives

• Establish taxonomy of email messages and their
  life-cycle
• Establish time attributes of email messages