Conference PaperPDF Available

What a to-do: studies of task management towards the design of a personal task list manager

Authors:

Abstract and Figures

This paper reports on the results of studies of task management to support the design of a task list manager. We examined the media used to record and organize to-dos and tracked how tasks are completed over time. Our work shows that, contrary to popular wisdom, people are not poor at prioritizing. Rather, they have well-honed strategies for tackling particular task management challenges. By illustrating what factors influence task completion and how representations function to support task management, we hope to provide a strong foundation for the design of a personal to-do list manager. We also present some preliminary efforts in this direction.
Content may be subject to copyright.
What a To-Do: Studies of Task Management Towards
the Design of a Personal Task List Manager
Victoria Bellotti, Brinda Dalal, Nathaniel Good, Peter Flynn, Daniel G. Bobrow
and Nicolas Ducheneaut
Palo Alto Research Center
3333 Coyote Hill Road,
Palo Alto, CA 94110
bellotti@parc.com
650 812 4000
School of Information
Management & Systems
University of California, Berkeley,
CA 94720
ngood@sims.berkeley.edu
Cornell University
Ithaca, NY 14853
psf2@cornell.edu
ABSTRACT
This paper reports on the results of studies of task
management to support the design of a task list manager. We
examined the media used to record and organize to-dos and
tracked how tasks are completed over time. Our work shows
that, contrary to popular wisdom, people are not poor at
prioritizing. Rather, they have well-honed strategies for
tackling particular task management challenges. By
illustrating what factors influence task completion and how
representations function to support task management, we
hope to provide a strong foundation for the design of a
personal to-do list manager. We also present some
preliminary efforts in this direction.
Categories & Subject Descriptors: H.4.1 [Information
Systems Applications]: Office Automation – time
management; H.5.1 [Information Interfaces and
Presentation]: Multimedia Information Systems –
evaluation/methodology; Keywords: ethnography.
General Terms: Design, Human Factors.
Keywords: Task management, to-dos, ethnography.
INTRODUCTION
A major initiative being considered at DARPA is directed
towards developing cognitive systems that can support busy
professionals in government or military roles in managing
and even performing office and military tasks. These
‘cognitive assistants’ will be capable of reasoning and
learning and will be aware of and able to explain their own
behavior as well as accepting direction from users [8]. One of
the possible embodiments of this type of system is a task list
manager system (TLM) that could help users manage and
execute their to-dos. Such a system would:
Capture the person’s daily tasks.
Plan and execute simple actions.
Prioritize, manage, and reason about tasks.
Learn to improve by being told, observing the user,
asking questions, and reflection.
Record notes, action items and ideas.
Answer questions and offer advice and assist in
planning and problem solving.
As part of this initiative, an initial effort at our laboratory was
undertaken to understand natural practices of task
management and types and quantities of tasks taken on by
people similar to prospective users of DARPA technology,
namely busy professionals and managers. In particular we
sought to discover what kinds of task management demands
might be supported by a TLM.
Prior Work
There are a number of best sellers and tools available on how
to organize one’s time and prioritize work [e.g., 1, 7]. The
market for these resources seems to thrive on the fact that
many people worry about whether they are prioritizing and
meeting their many obligations effectively [4].
Personal information management (PIM; organization, note
taking, reminding and calendaring) has also been examined
in the HCI literature, but mainly focusing on the problem of
organizing documents, files and notes for the purpose of
reminding and efficient retrieval, rather than task
management [2, 6, 9, 12, 13, 14]. There are a number of
studies on how people use calendars in practice (for far more
than just event scheduling), [5, 16, 17]. But this literature
focuses only on a single resource that mainly serves time
management needs. And the many readings available on
cognition, planning and task execution in the classical
psychological literature [10, 15] have little to say about task
management and planning in work practice.
Distributed cognition analyses [11, 18] comes closest to the
kind of analysis we seek here showing that external resources
are critical in performing complex tasks, but the literature
does not look at how external representations function to help
their creators assess the current state, extent and priority of
many tasks to be completed. The focus has been on resources
supporting task execution rather than articulation work (the
Permission to make digital or hard copies of all or part of this work for
p
ersonal or classroom use is granted without fee provided that copies are
not made or distributed for profit or commercial advantage and that copies
b
ear this notice and the full citation on the first page. To copy otherwise,
or republish, to post on servers or to redistribute to lists, requires prio
r
specific permission and/or a fee.
CHI 2004, April 24–29, 2004, Vienna, Austria.
Copyright 2004 ACM 1-58113-702-8/04/0004…$5.00.
CHI 2004 ׀ Paper 24-29 April ׀ Vienna, Austria
Volume 6, Number 1
735
work required to plan and organize work) [19]. And task
management from an articulation standpoint goes beyond
simply organizing physical and virtual collections and
putting events in a calendar.
Recent work has begun to tackle the challenge of general task
management at it plays out in email [3, 20, 23]. Given that
many PC users are overwhelmed by the number of tasks
handled through email, this trend is not surprising. This work
suggests that any successful task management resource must
be closely integrated with email functions. But since email is
unlikely to be the entire story, we felt we needed to look at
task management practices more broadly.
In the remainder of this paper we briefly discuss results of a
“snapshot” pilot study and then focus on a longer-term study
of task management representations and practices. We also
discuss our embodiment of some of our fieldwork-driven
ideas in a preliminary design. In our reporting, all people,
project and organization names have been changed or
obscured to protect the privacy of our participants.
A SNAPSHOT STUDY OF TO-DO’S
Prior to beginning an in-depth study of task management, we
conducted a pilot study in which 3 administrative staff, 4
researchers and one manager were asked to show all the
resources where they kept to-dos and count all of the active
to-dos they were currently tracking in each one. Table 1
shows average numbers of to-dos and resources used to
represent them, based on 595 counted by the 8 participants.
We also captured explanations of each resource and 79
detailed explanations of examples of to-dos, taking one or
two from each resource each person showed us.
Some key properties of effective to-dos quickly emerged, a
selection of which are discussed in the following:
To-dos are made expending minimal effort, so most of them
do not describe the task, they typically are only elaborated
enough to provide a salient cue. For example one to-do was
some text on a pad of paper; ‘Joe the attorney.’ The
explanation was, “A reminder to send him mail. I think I was
supposed to ask him about this […] lawsuit. I can't
remember.” Interestingly, to-do text is often not grammatical
as in, “Send Mother’s Day” or even “Beth blah blah”. The
cue is so minimal that it is only effective for a limited period
of time while the task stays in memory. In some cases an
item such as a book or a pile serves as enough of a cue to
recall the task, without creating a note.
Only a minority of to-do reminders appear in lists. We found
only 14% of the to-dos we counted were in a list.
To-dos are used in multiple ways. Sometimes they are part of
a list that provides a sense of the amount of work to do.
Sometimes they are resources supporting consultation,
linking to work objects, or are work objects themselves,
displaying state as well as to-do-ness.
Many to-dos are prompts placed in-the-way in anticipation of
a routine practice that will occur at the right moment for the
to-do to be discovered. For example, “When I go to grab my
bag to go home, I'll go, ‘I must take that [object next to bag]
home.’” Emails left in the inbox in particular serve this
prompting function during email-centered work.
To-dos may be represented at any level of abstraction or
detail (see Figure 1). We saw one to-do (not shown) that was
detailed enough to support preparation of some slides, but
other “give presentation” to-dos only referred to the subject
of the presentation as shown in Figure 1 (top right).
To-dos don’t all get done. People procrastinate about some,
and deem some of lower importance. Two participants
mentioned deliberately keeping low priority to-dos in an
electronic form that would be lost if an application or
machine crashed. An effective way of reducing the task list!
Implications for Task List Manager Design
There were some clear design implications from the above
findings. First, of our participants’ 70 or so to-dos, about half
are already online, even without the incentive of smart to-do
management support. The main resources are email and the
electronic calendar, but these have weaknesses [3, 20, 5, 17].
Thus, 50% of what is going on may be tractable to
improvement with a TLM if it can capture this activity, and
possibly more if users are motivated to move more to-dos
online because of the benefits of system support. Many
challenges are apparent however. A TLM must offer:
Diverse ways to view and manipulate to-dos to emulate
advantages of existing resources, going beyond lists.
The in-the-way property, e.g., by becoming the habitual
place for routine activities where reminders might be
encountered.
Instantly on, to support quick and easy input and clear
Figure 1. To-dos showing various levels detail from telling
someone (Name) to take an SGBS course (bottom right) to
preparing a presentation (Subject) presentation (top left).
Subject
N
ame
Per Person Ave Median %
No. of resources used to represent to-dos 11.25 11.5
No. of active to-dos 74.4 65
In email 26.6 20 35.8
In online calendar 8.6 9 11.6
In paper list or paper notepad 7.2 0 9.7
In online folders 3.5 1.5 4.7
In online special purpose to-do list 3 0 4.0
In PDA calendar and list combined 2 0 2.7
In daytimer/bound notebook/planner 1.6 0 2.2
Table 1. Number of to-do resources and active to-dos per
person. Online resources are shaded.
CHI 2004 ׀ Paper 24-29 April ׀ Vienna, Austria
Volume 6, Number 1
736
visualization. PDA’s are often abandoned due to slow
laborious input and attenuated output [6].
No formal task description, categorization or
decomposition required from users, and any level of
abstraction must be allowed for atomic task entries.
A mechanism for handling stale to-dos of low
importance that are diminishingly likely to ever get done
but have not been explicitly deleted.
LONGER-TERM STUDY OF TASK MANAGEMENT
The snapshot study, while useful, left many important
questions for a TLM unanswered, for example:
What help do people need with prioritization?
What factors drive people to prioritize and complete
tasks such that a TLM can propose appropriate action?
What is the lifespan of a useful to-do?
What kinds of task management resources are
appropriate for different challenges?
In order to address these questions, a longer-term study was
undertaken to capture and track a large number of to-dos.
Due to the heavy time-commitments required, only 7
participants were engaged. Participants were specifically
selected for having highly multitasking and diverse work
regimes but none were using any task-management
techniques such as the Franklin Covey system:
M1: Manager of between 5 and 9 research staff in our
organization. Reviews intellectual property (IP), plans and
conducts research and obtains external funding.
M2: Manages 15 to 20 researchers in our organization.
Tracks IP and does business development.
Prof: Professor and co-director of a research laboratory in a
university and manager of 5 to 20 people. Conducts and
manages research, and obtains funding.
SPM: Senior product manager signing up and managing
200 tour operators who sell their products through his
online company. Works in a small office in the USA while
his head office is in London, UK.
DDM: Director of Development and Marketing for a
charity. Writes grant proposals, liaises with donors and
supervises 3 to 4 staff members.
DM: District Manager of ten stores in a chain of retail food
and beverage stores. Visits stores, supervises store
managers (20 people), tracks and develops business.
SAM: Sales account manager in a large office-supplies
retail and wholesale company. Has approximately 300
ongoing and prospective accounts.
The data collection method (executed at all of the
participants’ places of work) was as follows:
Preliminary 2-hour background information interview.
Four 1-hour, weekly task-tracking interviews (referred to
as TT1, TT2, TT3 and TT4 respectively) in which
(usually) 10 to-dos that might be done by next week
were elicited, discussed and rated for importance and
urgency on a scales of 1 to 5. All to-dos from each week
were followed up at the next interview. Participants were
prompted not to focus on a single to-do resource or only
important and urgent to-dos.
One day of shadowing in which the participant was
observed doing normal daily tasks.
Final 1-hour interview to answer remaining questions
and follow up on the final disposition of all to-dos.
All sessions were video taped and transcribed.
All 287 to-dos were coded with a value of ‘yes’ or ‘no’ for
37 codes such as ‘done,’ ‘on-hold,’ ‘common,’
‘discretionary.’ Cross-coder reliability was obtained for a
subset of 50 tasks (50x37=1850 codings) with agreement of
92%. Correlations between coded factors were obtained with
both Pearson’s R and Spearman Correlation tests.
Task Tracking Findings
A minority, 104, of our tracked 287 tasks represented 45
tasks that were carried over one or more weeks. Thus we
have 228 unique tasks.
We were unable to observe the tasks we were tracking but we
got estimates for completion times ranging from 30 seconds
to 5 days with 63% being between 10 minutes and 4 hours.
Recollections were the same as the predictions in only 25%
of cases. In 40% of cases, recalled times were shorter,
averaging 77% of the duration of predicted times. In 35% of
cases they were longer, averaging 130% of the duration of
the predicted time. This could be because tasks took more or
less time than planned, and/or because people are poor at
estimating and recalling time required to do a task. In either
case (and this was confirmed informally in discussion with
our participant) predictions of task time are often inaccurate
(sometimes grossly so). Thus TLM users should not be relied
upon for precise time planning.
Medium of reminder No. % of
R
I U %Done
in 1 wk
%Dealt
by end
Email 88 34.8 3.2 2.9 68 82
Formal
p
a
p
er to-do 61 24.1 4.3 3.2 52 77
Print out
(
s
)
28 11.1 3.1 2.7 64 79
E-Calendar Entr
y
28 11.1 3.3 3.2 61 82
Index car
d
17 6.7 3.6 3.4 94 94
Pa
p
er sheet/
p
ad note 13 5.1 2.9 2.8 69 77
Ob
j
ect
;
book/ba
g
/etc
,
9 3.6 2.7 2.7 56 56
Note in noteboo
k
4 1.6 3.3 2.3 100 100
Vertical folders 3 1.2 3.2 3.7 100 100
Voicemail 1 0.4 4.5 4.5 100 100
O
en window on PC 1 0.4 4.0 3.0 100 100
In head
(
no remindr
)
34 0 3.8 3.3 88 85
Totals or averages 287 100 3.5 3.0 68 81
Table 2. Media used to record to-dos, their prevalence as a
percentage of all to-dos with reminders (R: 287-34=253),
average importance and urgency of to-dos per medium and
percentage of each type done in a week or by end of study.
Table 2 summarizes how the 287 to-dos were represented (in
fact 34 had no reminder representation) and shows that, on
average, participants did not offer only important (I) and
CHI 2004 ׀ Paper 24-29 April ׀ Vienna, Austria
Volume 6, Number 1
737
urgent (U) tasks to track. Urgency was not found to be
significantly correlated with any reminder medium.
The use of to-do lists was higher in this longer study (24%)
than in Table 1 (9% + 4.7%). Being on a list was positively
correlated (p<0.001) with importance and negatively
correlated with getting done in a week (p<0.001). Our data
thus confirms the idea that more important to-dos get onto
lists as they risk being overlooked while being delayed.
To-dos as email, print-outs, notes on paper or objects were
less important on average (p<0.05). Index cards were
positively correlated with getting done in a week, (but these
were used by only one person and so the ‘done’ factor is
confounded with ‘participant’). And objects were negatively
correlated with getting done in a week (P<0.05) suggesting
these object reminders often have low priority.
Having no reminder at all was significantly positively
correlated with completion in a week (p<0.01). However, it is
not correlated with importance, so it is not criticality that is
making these tasks easier to remember. We did get some
verbal reports that routine tasks tended not to be recorded as
to-dos because they are easily remembered habits.
By the end of the study, though, we found that there was no
statistical relationship between reminder medium and task
completion, except for object to-dos, which were
significantly less likely to be dealt with in the end (p<0.05).
Table 3 shows further significant (p<0.05; less than 5%)
chance of this data occurring randomly) correlates of task
completion within a week (see Table 3).
Factor Done if
high/yes
Done if
low/no
Sig p
level
Urgency (high/low) 93% 44% 0.001
Meeting (yes/no) 87% 66% 0.05
Importance (high/low) 81% 42% 0.00
Someone expecting (yes/no) 77% 53% 0.00
Other(s) involved; not mtg (yes/no) 82% 62% 0.00
Non-discretionary (yes/no) 79% 63% 0.02
Table 3. Factors related to task completion within a week.
Highly important and urgent tasks (rated 5 on a 5 point scale)
were twice as likely to get done as those rated low
importance or urgency (rated 1; note that these factors were
scalar, not binary like the rest). However, having a deadline
was not significant suggesting that people do not rate tasks as
important or urgent just because they have a time limit.
Tasks that were meetings or simply involved other people
were both more likely to get done. Someone else expecting
the task to be done was also highly significant. Whether or
not a task was self assigned or assigned by someone else was
not a factor, but discretionary tasks were less likely to be
completed than non-discretionary.
One of our most surprising findings is that our participants
seem to be prioritizing very competently. Only 3% of tasks
were dropped with no good reason. Since this finding flies in
the face of popular literature on task management (which
argues that people often fail to do important tasks due to poor
prioritization) we will examine it in more detail.
Of all 287 tasks tracked across our participants:
68% of all tasks were done in a week and 81% were
dealt with (as far as possible) by the final interview.
By the final interview, 79% of TT1, 81% of TT2, 83%
of TT3 and 80% of TT4 tasks were completed. Note that
the final interview was usually two weeks after TT4
explaining the higher completion rate than 68%.
The fact that there is no difference between older and newer
tasks being completed suggests that tasks planned for the
weeks ahead that are not completed after two weeks are
unlikely ever to get done. Qualitative data showed that to-dos
were sometimes held up by unplanned overload. For example
DDM admitted that she did not complete most of her to-dos
one week due to an unexpected and onerous task with a tight
deadline. However, all but one of these tasks were completed
by the following week.
A further 16% of tasks were not done by the final interview,
with this being a satisfactory outcome. In these cases,
participants gave us a very good reason for not completing
these tasks, for example (paraphrasing for brevity):
I don’t have to do this for several months (M2)
We don’t qualify for this funding cycle (DDM)
The customer wasn’t interested in the product (SAM)
The 3% of tasks dropped without good reason were always
either possible to do at some later point or were minor slip-
ups. We have no reason to suspect our participants were
dishonest about failure, as they showed no hesitation when
admitting to it (usually smiling or laughing).
So a surprising conclusion is that the problem with task
management is not failure to prioritize well. We would argue
rather that it is the effort that must go into making sure that
the important tasks get done, even if the unexpected occurs,
that is the real challenge. In previous research we found
people express great concern about this and frequently switch
between PIM resources to try to improve efficiency [4]. So
our next section discusses how professionals manage to be so
competent at prioritization. By learning from their success a
TLM can emulate and improve on the resources and
mechanisms they use.
We conclude this section, with the caveat that an 81% to-do
completion rate might be higher than the true completion
rate. Inevitably, participants chose the tasks that we tracked
and these may have been ones that were less likely to be
overlooked. However, as mentioned, to minimize this effect,
we encouraged participants to include many less obvious
tasks (hidden in piles, the email in-box etc).
CHI 2004 ׀ Paper 24-29 April ׀ Vienna, Austria
Volume 6, Number 1
738
Task Management Strategies
A number of task management strategies emerged that
seemed, not only to reflect personal preferences, but often to
be closely tied with the demands of the participants’
particular job pressures. We will discuss some of these in the
following sections (with the participants who used these
strategies shown in brackets):
Task Vistas (DDM, DM, M1)
Task vistas are fairly comprehensive lists used for planning,
ensuring that nothing (that could be forgotten if it were not
listed here) falls through the cracks. They reflect a desire to
be able to see all to-dos together on one page (all 3 listers in
this study used paper), hence our term ‘task vista.’ As DM
put it “I’ll shrink the font and change the margins to keep
everything all on one page. […] It just makes it more
digestible for me I think and I can see the connections….”
DDM and DM produced weekly updated task vistas with
systematic categories of tasks to assist in planning. Many
items reappeared week after week. DM’s list is particularly
interesting as she created many categories (see Figure 2). For
example she keeps things to talk to her manager about in one
place “Funnel to [blanked] /RP”, (bottom left) and things for
her support team in another (lower left). These persistent
categories reflect and support her routines and also break the
many disparate tasks that are a feature of her job into more
manageable chunks that she can plan around.
M1 showed us a different kind of ad hoc, occasional task
vista that he created when he felt overwhelmed, and
organized around projects, unlike the systematic lists. M1’s
lists were not updated, but simply replaced with a new list
with different structure when needed. He was managing
and/or participating in about 10 complex projects and
proposals during our study.
Informal Priority Lists (Prof, SAM)
Informal priority lists are selective task views that ensure
near-term execution of priority actions. These were not
organized and were jotted on small scraps of paper. Both
Prof and SAM created them at home for the coming work
day, daily in SAM’s case, and most days in Prof’s case. The
small size of the paper seems to be important for Prof as she
carries the list in a pocket to work with her.
State Tracking Resources (SPM SAM)
State tracking resources are needed when many similar (and
confusable) threads are on the go at once. They show what
actions were performed when and support looking back in
order to infer what must be done instead of explicitly
outlining what must be done. SPM and SAM both need to
keep track of a great many simultaneous ongoing threads of
activity. SPM has 200 or so prospective tour operators and
SAM has 300 or so leads and accounts. Since SPM is not
highly mobile, he uses a whiteboard with all of the accounts
listed with the last action and date usually recorded. SAM
needs a mobile solution as he is often out meeting clients and
prospective clients face-to-face, so he staples a business card
to an index card and write dates and notes on status and other
useful information on that.
Time Management (M1, M2, Prof, DDM, DM)
Time management resources are needed when time becomes
a scarce or inflexible commodity. The five participants who
seemed to pay most attention to timing were, unsurprisingly
the ones who were most subject to deadline pressures
(accounting for 93% of the deadline task codes). We saw
three strategies for time management:
The first is to make sure that things get noticed later,
when no time is available now. This is done by placing a
to-do ‘event’ in the calendar at some future date.
The second is to carefully preserve time for critical tasks
in a busy schedule where appointments could fill up the
days. Thus time would be blocked out in the calendar for
things that were not events. As Prof put it “…it’s a time
slot that I wanted to protect.”
The third, used by DM, is to categorize some tasks on
her list into either under 15 minutes or 30 minutes to 1
hour or more. She had around a dozen in the former and
half that number in the latter category, which she stated
made her task list seem far less daunting.
Value Extension (M1, Prof, SAM)
Value extension is the practice of making one task serve
multiple goals and further, to choose tasks to maximize this
effect. As M1 and Prof put it:
M1: “I spent some time talking to the people who are going
to commercialize it [software]. […] Thinking about how we
could actually use this for what we want to do. What can I do
to make it easy for those [interns] to try and do things with
that [software]. […] what are the small collaterals to make it
work in these different contexts.”
Fi
g
ure 2. DM’s to-do list: A s
y
stematic task vista.
CHI 2004 ׀ Paper 24-29 April ׀ Vienna, Austria
Volume 6, Number 1
739
Prof: “… we’re actually leveraging this work across a
number of projects.”
When shadowing SAM, we also noticed him stopping by
offices opportunistically on his way to clients’ offices to see
if he could drum up new business; having two clients close
together obviously saves on travel time.
Key Phenomena
Among our findings a number of phenomena, that are not
task management resources, are nonetheless key in
determining the outcomes of critical tasks:
Social Relationships: The Network Effect
People are highly sensitive to who is important to them in
their network when assigned tasks or getting requests:
SPM: “There are some people […] there [in London] that
work very, very hard for me, and if I have a choice, if I have
time to answer only one email, I'm going to answer that
person's email first because I know that they are going to act
on [it…], right away. They'll be appreciative.”
Another phenomenon is that when one person fails to
complete a task that others are waiting on, the network
compensates (a reminder or a demand materializes). And
when a person consistently fails, their colleagues develop
compensation strategies:
Prof: “[…] one thing I try really hard to do is get things off
my plate, so I ask for confirmation from unreliable resources,
‘I transferred this to-do’ to you. Can you confirm that you
have it?’ And then, if I don't trust them, I ask them to get back
to me on date x and date y.”
Thus, the implications of failure to complete a task are
nuanced since they depend on the nature of relationships
between people. And our participants often mentioned that
tasks that were important to other people were not
necessarily important to them. Whether they completed the
task or not depended on the relationship between them.
Working Away from the Desk
Among our participants, only SPM spends nearly all day in
his office. With the others, we saw various instances where
task management away from the desk was crucial. Many
tasks were created away from the desk through extensive
interaction in meetings and customer visits. For example,
during our shadowing, DDM and her marketing manager
created many actions for each of them in one meeting.
Rhythms and Routines
As with Tolmie et al. in their study of domestic life [21], we
noticed tasks had temporal and procedural regularities. One
kind of regularity is that a certain kind of task is likely to take
place at a certain time—of day, day of the week, time of year,
etc. For example, everyone begins the day by dealing with
email and/or voicemail. And DM generally visits particular
stores on the same day each week. Another kind of regularity
is that some tasks follow a pattern, which has temporal and
action-based consistency. For example: SAM makes a cold
call; if the client is interested, he follows up with a visit and
product information; the client (hopefully) opens an account;
SAM waits for some period and then goes online to inspect
activity in the account to see if a follow up is needed to
uncover any problems.
Implications for Task List Manager Design
Our findings have a number of implications for a TLM:
A TLM should support the viewing of entire task vistas, but
also allow different perspectives for different kinds of
planning. By this we mean a TLM must emulate the
properties of the different kinds of to-do lists we have
observed. Such views must be able to be sort and filter
activity to show a day or, week or project scope. It should
allow top priority items to be made apparent or items with
similar properties to be viewed together, e.g., items that can
be executed quickly, or those to be completed with a
particular person. As noted, DM has special categories for
both these types of items on her to-do list.
Task histories and state should be captured for the purpose
of being able to determine information such as, when the last
action occurred, what the current state is, who is responsible
for the next action and so on. This capacity would support
users who, like SPM and SAM, have many tasks ongoing at
once.
Time constraints should be captured such that decisions
about workload and scheduling can be made with a clear idea
of how the overall picture will be affected. Combining these
with knowledge about rhythms and routines could allow for
visualizations of truly free time, as opposed to unscheduled
time and support time and interval sensitive re-prioritizations
of a user’s tasks. However, a TLM should not rely on users
making accurate predictions about the time required to
complete their tasks. Indeed it should not require tasks to be
completed at all. However, if a TLM is to succeed at
representing time commitments, then users must be
encouraged to enter many tasks that they might not normally
represent as to-dos, simply to capture their time
commitments. This suggests that a TLM must offer more
benefits than just reminding for entering tasks, such as using
list entries as organizational or time management resources
or to launch the related tasks themselves.
The properties of tasks must be modeled in such a way that it
is possible to practice value extension more explicitly. For
example, making the location of a task explicit (as in the
Llamagraphics task management tool Life Balance™) allows
multiple tasks to be planned retrieved and executed at the
appropriate location, requiring only one excursion.
Social relations should be captured and modeled, perhaps
using information such as email responsiveness images (time
taken to respond to specific people [22]), which seem
predictive of the importance of social relationships. Doing so
can to help prioritize tasks such that those requested by
known valued colleagues are pushed to the top of the stack.
CHI 2004 ׀ Paper 24-29 April ׀ Vienna, Austria
Volume 6, Number 1
740
Fi
g
ure 3. TaskVista: a
p
ersonal
task list mana
g
er com
p
onent.
Finally, a TLM must support the capture of notes and task
lists away from the desk in order to be effective.
TOWARDS DESIGN FOR A TLM
During the period when we were intensively engaged with
fieldwork and developing our ideas about task management,
we felt it was also important to keep ourselves grounded in
solving design problems relevant to support for task
management. So we designed and prototyped an early
component of a TLM system that addresses some (but not
all) of our findings. This is described in the following
section. To assist the reader in linking its design features to
our ethnographic findings, we have italicized the properties
that our research indicates are likely to be valuable.
TaskVista
TaskVista (see Figure 3)
is a lightweight resource
for collecting and listing
to-dos and conveniently
launching tasks from
them. Unlike existing
tools such as the
Microsoft Outlook™ Task
List, it is compact and sits
on the desktop (or PDA)
and is in-the-way but not
obtrusive when users
reach transition points
between tasks (e.g., when
they switch applications).
It is a comprehensive to-
do list that easily handles
a realistic number of
active to-dos. Old to-dos
do not need to be deleted
but are filtered out of sight
when they become
defunct or are done, to
avoid clutter.
Users can easily create a new to-do by typing one in, or,
dragging an item (e.g. a file or email) into the list. The
(editable) title defaults to the subject or title of a dragged-in
item to reduce the user’s need to type. Additional items such
as notes, documents, etc., can be dragged in to a to-do, so the
to-do becomes resource for saving content and launching
activity on the task, like a pile or folder. But unlike a pile or
folder, a to-do has computational properties that support
task management. For example, each to-do has the property
of importance that determines its priority or position in the
list when the list is sorted for importance. Users can change
importance easily by dragging a to-do up and down the list.
A to-do can also have time constraints (e.g. a duration as in
DM’s list or Prof’s calendar time-slots, or a deadline). Green
‘warning bars’ turning red (also used in [3]) are a salient
visualization to cue users of the urgency of approaching
deadlines.
TaskVista also has a pop-out visualization (it slides out to the
right) that shows tasks in a temporal view laid out as bars in a
Gantt chart across a horizontal timeline to support time
management. Users can see which tasks coincide and can
directly manipulate date and time information.
Since we found that relationships are an important factor in
task management, to-dos also have the property of
participants. These are names extracted from sender and
recipient information in email messages or from document
content or properties and can be matched with details in a
contact list (which the user can mark up to record who is
important; this information could also be inferred from email
response profiles [22]). TaskVista provides a contact widget
(also used in [3]) for each task, so users can easily open a
message to all participants with two mouse-clicks or select a
subset one by one and then open a message to them.
To-dos can have other properties that, for example, show
location, task or participant dependencies, and whether they
are a project Properties do not need to be specified up-front,
and can change over time. The more information TaskVista
has, the better job it can do at what we call ‘smart’
prioritization. The user can elect to have the system prioritize
tasks, for example, because they are non-discretionary,
because the participants are important, or because a deadline
is approaching. It does so simply by pushing them up the list
in a ‘smart sort’ view. The list can be sorted or filtered based
on single or multiple properties to support different kinds of
task management activities.
Finally, users do not have to decide if a to-do is a simple task
or an entire project. They can specify to-dos at any level of
abstraction and turn tasks into projects (encompassing other
tasks) at any time depending on needs.
TaskVista was developed quickly in C# to allow us to
conduct a very early informal evaluation with 9 volunteers.
Users were guided through 10 task management exercises
based on a real experience of one of the authors who was
asked to insert an important and urgent presentation into an
already crowded schedule. This required creating a new task
from an email invitation to give a presentation, rating it as
important, mailing people involved in existing tasks to notify
them that one is now too busy to work on them, etc. This
effort, even with a buggy prototype, provided a great deal of
valuable feedback about how to refine our ideas for a TLM
towards something we feel would be truly effective.
Overall, even with an early prototype, users were positive
about many of its features, (rating them on average 3.94 in a
scale from 1 = hate-it to 5 = love-it; standard deviation
(stdev) was 0.84). The most popular features are the drag and
drop creation of to-dos (4.67 on our scale, stdev 0.71)
viewing the list on the desktop (4.56, stdev 0.73), the ability
to sort by importance and deadlines (4.2, stdev 0.83) and the
fact that it’s easy to get at task participants (4.4, stdev 0.73).
CHI 2004 ׀ Paper 24-29 April ׀ Vienna, Austria
Volume 6, Number 1
741
The temporal view has so far been the least well received
feature (3.0, stdev 0.71); it seems somewhat unnatural for
those not used to looking at Gantt charts, even though we
know that this is an accepted type of view for planning
complex projects.
A critical challenge and valuable feedback for DARPA is
that most users disliked the idea of explicitly adding a lot of
metadata such as participants, planned start and end times, or
tasks dependencies. Thus, the design of a TLM must
encompass lightweight or automatic mechanisms to capture
or infer such information. Fortunately one of the main thrusts
of the DARPA program is to create just such mechanisms,
which will incorporate reasoning and learning to lighten the
load of users in prioritizing and planning.
CONCLUSIONS
We propose that the principal problem of task management is
not poor prioritization, but the effort it requires and have
outlined resources and methods people use that help ensure
they are effective at this. We designed TaskVista as a tool to
reduce this effort. It is faithful to field-derived insights into
what factors and resources relate to task management.
Designing and evaluating it early, even as we conducted our
fieldwork, helped us maintain our focus as ethnographers in
generating design recommendations. Our early evaluation of
a TLM prototype confirmed our ethnographically derived
requirements for an interactionally lightweight tool with
intuitive visualizations and the capacity to work with
underspecified and arbitrarily abstract content as it naturally
occurs in task management practice.
ACKNOWLEDGMENTS
This work was supported in part by DARPA/IPTO under SPAWAR
Contract N66001-03-C-8010. The views and conclusions contained
in this document are those of the authors and should not be
interpreted as representing the official policies, either expressly or
implied of the Defense Advanced Research Projects Agency, or the
U.S. Government.
REFERENCES
1. Allen, D. Getting Things Done. Viking (2001).
2. Barreau, D. & Nardi, B. Finding and Reminding: File
Organization from the Desktop. SIGCHI Bulletin, 27, 3
(1995), 39-43.
3. Bellotti, V., Ducheneaut, N., Howard, M., & Smith, I.
Taking Email to Task: The Design and Evaluation of a
Task Management Centered Email Tool. In Proc. CHI
2003, CHI Letters 5 (1). 297-304.
4. Bellotti, V. & Smith, I. Informing the Design of an
Information Management System with Iterative
Fieldwork. Proc. DIS’00, Designing Interactive Systems.
ACM Press, (2000) 227-237.
5. Blandford, A. & Green, T. Group and Individual Time
Management Tools: What You Get is Not What You
Need. Personal and Ubiquitous Computing 5, 4 (2001)
213-230.
6. Campbell, C. & Maglio, P. Supporting Notable
Information in Office Work. Extended Abstracts CHI’03
Conference on Human Factors in Computing Systems,
ACM Press (2003), 902-903.
7. Covey, S. The Seven Habits of Highly Effective People,
Simon & Schuster (1989).
8. DARPA IPTO BAA 02-21: http://www.darpa.mil/-
ipto/Solicitations/open/02-21_PIP.htm
9. Hayes, G., Pierce, J. & Abowd, G. Practices for Capturing
Short Important Thoughts. Extended Abstracts CHI’03
Conference on Human Factors in Computing Systems,
ACM Press (2003), 904-905.
10. Hoc, J. Cognitive Psychology of Planning. Academic
Press (1988).
11. Hutchins, E. Cognition in the Wild. Cambridge, Mass:
MIT Press (1995).
12. Jones, S. & Thomas, P. Empirical Assessment of
Individuals’ ‘Personal Information Management
Systems’. Behaviour and Information Technology, 16.3,
(1997) 158-160.
13. Lansdale M. The Psychology of Information
Management. Applied Ergonomics, 19, (1988), 55-66.
14. Malone, T. How Do People Organize Their Desktops?
Implications for the Design of Office Information
Systems, ACM Transactions on Office Information
Systems, 1, (1983) 99-112.
15. Miller, G., Galanter, E. & Pribram, K. Plans and the
Structure of Behavior. Holt, Reinhart & Winston Inc.
(1960).
16. Palen, L. Social, Individual & Technological Issues for
Groupware Calendar Systems. Proc. CHI’99 Conference
on Human Factors in Computing Systems, ACM Press
(1999) 17-24.
17. Payne, S. Understanding Calendar Use. Human-
Computer Interaction, 8 (1993) 83-100.
18. Scaife, M. & Rogers, Y. External Cognition: How Do
Graphical Representations Work? International Journal
of Human-Computer Studies, 45 (1996) 185-213.
19. Strauss, A. Work and the Division of Labor. The
Sociological Quarterly, 26,1, (1985), 1-19.
20. Takkinen, J. & Shahmehri, N. Task-Oriented
Restructuring of an Application Domain: A Multi-Agent
Architecture for Doing Things in Internet E-Mail," Proc.
32nd Hawaii International Conference on System
Sciences (1999).
21. Tolmie, P., Pycock, J. Diggens, T., MacLean, A. &
Karsenty, A. Ubiquity. Unremarkable Computing. Proc
CHI 2002, CHI Letters 4(1).
22. Tyler J. & Tang J. Email Responsiveness and Rhythms:
Study and Design Implications. Proc. ECSCW’03
European Conference on Computer Supported
Cooperative Work. Kluwer (2003) 238-258.
23. Whittaker, S., & Sidner, C. Email Overload: Exploring
Personal Information Management of Email. In Proc.
CHI’96, Conference on Human Factors in Computing
Systems, ACM Press (1996) 276-283.
CHI 2004 ׀ Paper 24-29 April ׀ Vienna, Austria
Volume 6, Number 1
742
... Task management strategies range from scribbled sticky notes on refrigerators to complex group collaboration platforms such as Microsoft Planner, and everything in between. Digital tools for task management support are prevalent in a range of applications including electronic mail (Bellotti et al., 2003), to-do applications (Bellotti et al., 2004) and digital assistants (Graus et al., 2016), and for a range of user scenarios such as contextual reminders (Kamar and Horvitz, 2011), task duration estimation (White and Hassan Awadallah, 2019), and complex task decomposition (Zhang et al., 2021). Many of these applications record and utilize a number of signals such as the text of a task, the time it was created, it's importance, due-date, and who it is assigned to, in order to build intelligent solutions for the user. ...
... Two types of context that are especially salient are location and time. These signals are readily accessible to systems and have been studied in previous work on contextual understanding (Graus et al., 2016;Bellotti et al., 2004;Benetka et al., 2019), recommendation (Zhuang et al., 2011;Yao et al., 2015;Zeng et al., 2016) and reminders (Kamar and Horvitz, 2011). However, there continues to remain a disconnect between logged contextual signals and actions taken in the real world, since users often record completion of tasks at a later time and a different location (Zhang et al., 2022). ...
... Even before the existence of smaller computing devices such as cellphones that allow location-aware computing, actions as simple as setting an alarm or time-based reminder make it abundantly clear that there is a strong need for temporal context in successful task management. Bellotti et al. (2004) point out that co-location of tasks minimizes the need for multiple excursions for a user and that certain kinds of tasks show effects of periodicity, in terms of certain days, weeks and time of year. Graus et al. (2016) study reminder data from the Cortana personal assistant. ...
Preprint
Full-text available
Tasks are a fundamental unit of work in the daily lives of people, who are increasingly using digital means to keep track of, organize, triage and act on them. These digital tools -- such as task management applications -- provide a unique opportunity to study and understand tasks and their connection to the real world, and through intelligent assistance, help people be more productive. By logging signals such as text, timestamp information, and social connectivity graphs, an increasingly rich and detailed picture of how tasks are created and organized, what makes them important, and who acts on them, can be progressively developed. Yet the context around actual task completion remains fuzzy, due to the basic disconnect between actions taken in the real world and telemetry recorded in the digital world. Thus, in this paper we compile and release a novel, real-life, large-scale dataset called MS-LaTTE that captures two core aspects of the context surrounding task completion: location and time. We describe our annotation framework and conduct a number of analyses on the data that were collected, demonstrating that it captures intuitive contextual properties for common tasks. Finally, we test the dataset on the two problems of predicting spatial and temporal task co-occurrence, concluding that predictors for co-location and co-time are both learnable, with a BERT fine-tuned model outperforming several other baselines. The MS-LaTTE dataset provides an opportunity to tackle many new modeling challenges in contextual task understanding and we hope that its release will spur future research in task intelligence more broadly.
... They involve different activities, have different constraints, and take different amounts of time to complete. Users of task management applications would benefit from assistance with many aspects of task management, especially task planning [20] and prioritization [128]. There has been recent progress in task intelligence, in areas such as discovering digital assistant capabilities [176], estimating task durations [181], and automatically tracking task status [182]. ...
... Detecting task completion can be straightforward for some tasks, e.g., transactional tasks, but complex for others, e.g., learning tasks [183]. Progress can be tracked using dedicated tools [20] or inferred [182]. Recent research has built benchmarks for measuring task progress in digital assistants [104]. ...
Preprint
Full-text available
The importance of tasks in information retrieval (IR) has been long argued for, addressed in different ways, often ignored, and frequently revisited. For decades, scholars made a case for the role that a user's task plays in how and why that user engages in search and what a search system should do to assist. But for the most part, the IR community has been too focused on query processing and assuming a search task to be a collection of user queries, often ignoring if or how such an assumption addresses the users accomplishing their tasks. With emerging areas of conversational agents and proactive IR, understanding and addressing users' tasks has become more important than ever before. In this paper, we provide various perspectives on where the state-of-the-art is with regard to tasks in IR, what are some of the bottlenecks in deriving and using task information, and how do we go forward from here. In addition to covering relevant literature, the paper provides a synthesis of historical and current perspectives on understanding, extracting, and addressing task-focused search. To ground ongoing and future research in this area, we present a new framing device for tasks using a tree-like structure and various moves on that structure that allow different interpretations and applications. Presented as a combination of synthesis of ideas and past works, proposals for future research, and our perspectives on technical, social, and ethical considerations, this paper is meant to help revitalize the interest and future work in task-based IR.
... to re-find them later [2,3]. PIM research studies the collection and retrieval of personal pictures [4,5], websites [6,7], email [e.g., 8,9], personal health records [10,11], task management [12,13], and file documents (for literature review, see [14]). ...
Article
Full-text available
Streaming music applications allow their users to listen to music that they did not collect, making music collecting voluntary for the first time in the history of music consumption. Using a questionnaire (n = 370), we aimed to measure streaming collection size and its relations to other variables. Results showed a large variability in streaming collection size, with a median of only 200 songs. Furthermore, most participants drastically reduced their collection size when moving to streaming technology. Our results nevertheless found extensive evidence for the merits of streaming collections, including (a) collection size positively correlated with both listening hours and listening enjoyment; (b) participants wanted to listen to their favorite music for most of their streaming listening time; (c) young participants collected almost twice as many streaming songs as older participants, indicating that collecting is not an old habit; and (d) participants who enjoyed listening to records/CDs more than streaming were found to have previously been avid collectors who had abandoned collecting for streaming applications. In this paper, we discuss why regardless of these merits, the large majority of our participants reduced their collection size and conclude with suggestions about how the redesigning of streaming applications may encourage users to collect more songs.
... In this vein, HCI research investigated a variety of time-related practices, and, in so doing, they progressively attempted to go beyond the clock time. A first line of research that shows how HCI transitioned from an objective perspective on time to a more multifaceted conception refers to the time management practices (e.g., Schwarz, 1995, 1999;Bellotti et al., 2004). Traditionally, this kind of research has looked into developing technologies to improve efficiency of time usage, allowing for collective management of time and collaborative scheduling of tasks, through systems like Groupware (Palen, 1999) and Ambush (Mynatt and Tullio, 2001). ...
Article
The idea of time as an existential concern has found scarce attention in the Human-Computer Interaction (HCI) community. Still, time in our life is mostly experienced in connection with important existential matters. Life changes are emblematic examples of aspects of our existence that are fundamentally temporal. In this article, I explore how people experience the temporalities of “change”, by asking 16 participants to report on life changes they consider relevant and their temporal aspects. I found that temporal features like velocity, direction, and synchronization deeply impact the process of change, by affecting the participants’ opportunities to make sense of and act upon their own transformations. The study findings point to a conception of existential time that contributes to both HCI research on time and life changing events, by highlighting the relativity of temporal slowness and speediness, the malleability of time, and the fundamental openness of our time experiences.
... It is no secret that a large portion of our population struggle with task management [1]. According to a 2021 research, twenty five percent of people do not employ a task management system and simply work on "whatever seems the most important at the moment". ...
Conference Paper
It is no secret that a large portion of our population struggle with task management [1]. According to a 2021 research, twenty five percent of people do not employ a task management system and simply work on “whatever seems the most important at the moment”. Among the population that do use some sort of task management, the most popular form of managing personal tasks is through to do lists (33%) followed by using their email inbox (24%). Thus, I thought to combine these two most common methods by creating a to do list automatically generated from the email. We used the open sourced software natural language processing (NLP) to pick out the important sentences in the text and convert them into tasks for the to-do list. We used keywords such as “tomorrow”, “next”, “month”, etc combined with the date the email was sent to determine the “due date” of the to do list. Because the to-do list extracts information directly from the inbox without any participation from a human, unlike many other apps, this could be effectively used by those that do not check their email.
... From small start-ups to large technology corporations, modern information workplaces on Earth involve teams of people, often acting as experts of a particular information domain (e.g., writing, programming, data science), who cooperate toward an information-oriented end-goal through computers [52]. Today, information work is riddled by a myriad of challenges, such as task management [14], interrupted work [94], and burnout [31]. Modern research mitigates these challenges with various artificial intelligence techniques [50,74,137]. ...
Article
Full-text available
We present a design fiction, which is set in the near future as significant Mars habitation begins. Our goal in creating this fiction is to address current work-life issues on Earth and Mars in the future. With shelter-in-place measures, established norms of productivity and relaxation have been shaken. The fiction creates an opportunity to explore boundaries between work and life, which are changing with shelter-in-place and will continue to change. Our work includes two primary artifacts: (1) a propaganda recruitment poster and (2) a fictional narrative account. The former paints the work-life on Mars as heroic, fulfilling, and fun. The latter provides a contrast that depicts the lived experience of early Mars inhabitants. Our statement draws from our design fiction in order to reflect on the structure of work, stress identification and management, family and work-family communication, and the role of automation.
Article
Task management tools allow people to record, track, and manage task-related information across their work and personal contexts. As work contexts have shifted amid the COVID-19 pandemic, it has become important to understand how these tools are continuing or failing to support peoples’ work-related and personal needs. In this paper, we examine and probe practices for managing task-related information across the work-life boundary. We report findings from an online survey deployed to 150 information workers during Summer 2019 (i.e., pre-pandemic) and 70 information workers at the same organization during Summer 2020 (i.e., mid-pandemic). Across both survey cohorts, we characterize these cross-boundary task management practices, exploring the central role that physical and digital tools play in managing task-related information that arises at inopportune times. We conclude with a discussion of the opportunities and challenges for future productivity tools that aid people in managing task-related information across their personal and work contexts.
Article
People rely on task management applications and digital assistants to capture and track their tasks, and help with executing them. The burden of organizing and scheduling time for tasks continues to reside with users of these systems, despite the high cognitive load associated with these activities. Users stand to benefit greatly from a task management system capable of prioritizing their pending tasks, thus saving them time and effort. In this article, we make three main contributions. First, we propose the problem of task prioritization, formulating it as a ranking over a user’s pending tasks given a history of previous interactions with a task management system. Second, we perform an extensive analysis on the large-scale anonymized, de-identified logs of a popular task management application, deriving a dataset of grounded, real-world tasks from which to learn and evaluate our proposed system. We also identify patterns in how people record tasks as complete, which vary consistently with the nature of the task. Third, we propose a novel contextual deep learning solution capable of performing personalized task prioritization. In a battery of tests, we show that this approach outperforms several operational baselines and other sequential ranking models from previous work. Our findings have implications for understanding the ways people prioritize and manage tasks with digital tools, and in the design of support for users of task management applications.
Article
Applying lessons learned developing and deploying task management software.
Chapter
In the new economic landscape, companies wish to become agile and instantly react to occurring opportunities, and moreover wish to offer their clients excellent experiences, take advantage of the new technologies, and cutting costs. Thus, they need employees who show intelligent attitudes toward work. This paper is designed based on a literature review regarding interdisciplinary theories gravitating around the concept of innovating task management processes—activity management in the digital era. In this context, the authors have developed a visual tool to manage tasks within a project/job, allowing the development of studies focused on the stimulation and identification of employees showing an intelligent attitude toward work, essential for the company’s development, for the increase of digital business and for maintaining market competitive advantage.KeywordsTask ManagementVisualizationFast Syncing
Book
Taking a psychological perspective, this book examines the role of Human-Computer Interaction (HCI) in the field of mobile information systems research. The text begins by exploring the basic tenets of the HCI discipline, including the interdisciplinary nature of the subject, with increasing focus on the application of HCI methodologies and techniques in mobile HCI research. The author examines the research areas where HCI has not yet been fully exploited in relation to mobile information systems and services, such as broadening user groups and user acceptance of technology, and a range of guidelines and technical tips are provided to aid students looking to undertake their own mobile HCI-based research projects.
Article
This article is an interview study of calendar use and a cognitive analysis of the interactions between the design of calendars and the task of prospective remembering. The study and analysis are coordinated to present a general critique of current electronic calendar designs and to note opportunities for future designs. The interview data reveal continued use of paper calendars in a highly computerized setting. A key conclusion is that paper calendars support prospective remembering by promoting browsing of existing appointments during subsequent calendar keeping but that this advantage is compromised in current electronic designs. Other aspects of the interviews and the analyses address the representational limitations of both paper and electronic calendars. This research explores a methodology in which design implications of qualitative empirical data are understood and systematized through theoretical analyses of existing artifacts.
Article
This paper summarizes and synthesizes two independent studies of the ways users organize and find files on their computers. The first study (Barreau 1995) investigated information organization practices among users of DOS, Windows and OS/2. The second study (Nardi, Anderson and Erickson 1995), examined the finding and filing practices of Macintosh users. There were more similarities in the two studies than differences. Users in both studies (1) preferred location-based finding because of its crucial reminding function; (2) avoided elaborate filing schemes; (3) archived relatively little information; and (4) worked with three types of information: ephemeral, working and archived. A main difference between the study populations was that the Macintosh users used subdirectories to organize information and the DOS users did not.
Article
. A simple pilot study has been undertaken to investigate the use of 'personal information management technologies' (including both traditional and emerging electronic technologies) via a combination of nominal questionnaire and semi-structured interview schedules. The pilot study suggested that the number of users adopting computer-based personal information management technologies is low, as is the combination of computer-based technologies with traditional paper-based technologies in users' 'personal information management systems'.