Conference PaperPDF Available

Am I wasting my time organizing email?


Abstract and Figures

We all spend time every day looking for information in our email, yet we know little about this refinding process. Some users expend considerable preparatory effort creating complex folder structures to promote effective refinding. However modern email clients provide alternative opportunistic methods for access, such as search and threading, that promise to reduce the need to manually prepare. To compare these different refinding strategies, we instrumented a modern email client that supports search, folders, tagging and threading. We carried out a field study of 345 long-term users who conducted over 85,000 refinding actions. Our data support opportunistic access. People who create complex folders indeed rely on these for retrieval, but these preparatory behaviors are inefficient and do not improve retrieval success. In contrast, both search and threading promote more effective finding. We present design implications: current search-based clients ignore scrolling, the most prevalent refinding behavior, and threading approaches need to be extended.
Content may be subject to copyright.
Am I wasting my time organizing email?
A study of email refinding
Steve Whittaker, Tara Matthews, Julian Cerruti, Hernan Badenes, John Tang
IBM Research - Almaden
San Jose, California, USA
{sjwhitta, tlmatthe}, {jcerruti, hbadenes},
We all spend time every day looking for information in our
email, yet we know little about this refinding process. Some
users expend considerable preparatory effort creating
complex folder structures to promote effective refinding.
However modern email clients provide alternative
opportunistic methods for access, such as search and
threading, that promise to reduce the need to manually
prepare. To compare these different refinding strategies, we
instrumented a modern email client that supports search,
folders, tagging and threading. We carried out a field study
of 345 long-term users who conducted over 85,000
refinding actions. Our data support opportunistic access.
People who create complex folders indeed rely on these for
retrieval, but these preparatory behaviors are inefficient and
do not improve retrieval success. In contrast, both search
and threading promote more effective finding. We present
design implications: current search-based clients ignore
scrolling, the most prevalent refinding behavior, and
threading approaches need to be extended.
Email, refinding, management strategy, search,
conversation threading, folders, usage logging, field study,
ACM Classification Keywords
H5.3 Group and Organization Interfaces: Asynchronous
interaction, Web-based interaction.
The last few years have seen the emergence of many new
communication tools and media, including IM, status
updates, and twitter. Nevertheless, in work settings email is
still the most commonly used communication application
with reported estimates of 2.8 million emails sent per
second [15]. Despite people’s reliance on email,
fundamental aspects of its usage are still poorly understood.
This is especially surprising because email critically affects
productivity. People use email to manage everyday work
tasks, using the inbox as a task manager and their archives
for finding contacts and reference materials [2,7,23]. This
paper looks at an important, under-examined aspect of task
management, namely how people refind messages in email.
Refinding is important for task management because people
often defer acting on email. Dabbish et al. [7] show that
people defer responding to 37% of messages that need a
reply. Deferral occurs because people have insufficient time
to respond at once, or they need to gather input from
colleagues [2,23]. Refinding also occurs when people return
to older emails to access important contact details or
reference materials.
Prior work identifies two main types of email management
strategies that relate to different types of refinding
behaviors [13,23]. The first management strategy is
preparatory organization. Here the user deliberately creates
manual folder structures or tags that anticipate the context
of retrieval. Such preparation contrasts with opportunistic
management that shifts the burden to the time of retrieval.
Opportunistic refinding behaviors such as scrolling, sorting
or searching do not require preparatory efforts. Previous
research has noted the trade-offs between these
management strategies. Preparation requires effort, which
may not pay off, for example if folders do not match
retrieval requirements. But relying on opportunistic
methods can also compromise productivity. Active
foldering reduces the complexity of the inbox. Without
folders, important messages may be overlooked when huge
numbers of unorganized messages accumulate in an
overloaded inbox [2,7,23].
Choice of management strategy has important productivity
implications since preparatory strategies are costly to enact.
Other work has shown that people spend an average of 10%
of their total email time filing messages [3]. On average,
they create a new email folder every 5 days [5]. People
assume that such preparatory actions will expedite future
retrieval. However, we currently lack systematic data about
the extent to which these folders are actually used, because
none of these prior studies examined actual access
behaviors. Such access data would allow us to determine
whether time spent filing is time well spent. This is
important because prior work suggests that organization can
be maladaptive, with people creating many tiny ‘failed
folders’ or duplicate folders concerning the same topic [23].
Permission to make digital or hard copies of all or part of this work for
personal or classroom use is granted without fee provided that copies are
not made or distributed for profit or commercial advantage and that copies
bear 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 prior
specific permission and/or a fee.
CHI 2011, May 7–12, 2011, Vancouver, BC, Canada.
Copyright 2011 ACM 978-1-4503-0267-8/11/05....$10.00.
Another important reason for reexamining how people
manage and access email is the emergence of new search-
oriented clients such as Gmail [12]. Such clients assume the
benefits of the opportunistic approach as they do not
directly support folders. A second novel characteristic is
that they are thread-based. Building on much prior work on
email visualization [2,19,20], Gmail offers intrinsic
organization, where messages are automatically structured
into threaded conversations. Threads potentially help
people more easily access related messages. A thread-based
inbox view is also more compact, enabling users to see
more messages without scrolling, helping people who rely
on leaving messages in the inbox to serve as ‘todo’
reminders. We therefore examine the utility of these new
email client features by determining whether search and
threads are useful for retrieval.
We extend approaches used in prior work that tried to
identify email management strategies by analyzing single
snapshots of email mailboxes for their structural properties,
such as mailbox size, number of folders, and inbox size,
[2,11,13,23]. We also know that users are highly invested in
their management strategies [2,23] so it is important to
collect objective data about their efficacy. We therefore
logged actual daily access behaviors for 345 users enacting
over 85,000 refinding operations, and looked at how access
behavior relates to management strategy. Our method has
the benefit of capturing systematic, large-scale data about
refinding behaviors ‘in the wild’. It complements smaller-
scale observational studies of email organization [1,2,23],
and lab experiments that attempt to simulate refinding
(‘find the following emails from your mailbox’) [10].
Finally our study also extends the set of users studied.
Unlike prior work, only 2% of our users are researchers.
To apply this logging approach we needed to implement
and instrument a fully featured modern email client. Later,
we describe the client used to collect this data, which
supports efficient search, tags, and threading.
This paper looks at the main ways that people re-access
email information, comparing the success of preparatory vs.
opportunistic retrieval. We explore how two aspects of
refinding interrelate. On one level we wish to characterize
basic refinding behaviors to determine whether people
typically search, scroll, access messages from folders, or
sort when accessing emails. We also want to determine the
efficiency and success of these different behaviors, as well
as how behaviors interrelate. At the next level, we want to
examine the relationship between refinding behaviors and
people’s prior email management strategies, to determine
for example, whether people who have constructed complex
folder organizations are indeed more reliant on these at
retrieval. We therefore ask the following specific questions:
Access behaviors: What are people’s most common email
refinding behaviors, when provided with a modern client
that supports search, tagging, and threads, as well as
folders? Do people opportunistically refind emails by
scanning their inbox, searching, or sorting via header data?
Or instead do they use preparatory behaviors that exploit
pre-constructed organization in the form of folders or tags?
Also, what are the interrelations between behaviors? For
example, are there people who rely exclusively on search
and never use folders for access?
Relations between management strategy and access
behaviors: Does prior organizational strategy influence
actual retrieval? Are people who prepare for retrieval by
actively filing, more likely to use these folders for access?
In contrast, are people who make less effort to prepare for
retrieval more reliant on search, scanning, and sorting?
Impact of threads on access: Do threads affect people’s
access behaviors? Are people with heavily threaded emails
less reliant on folders for access?
Efficiency and success of management strategies and
access behaviors: We also wanted to know whether access
behaviors affect finding outcome. Which behaviors are
more efficient and which lead to more successful finding?
We might expect folder-access to be more successful than
search, as people have made deliberate efforts to organize
messages into specific memorable categories. On the other
hand, search may be more efficient as it might take users
longer to access complex folder hierarchies. Finally, are
people who create many folders more successful and
efficient at retrieval?
Studies of email use have documented how people use
email in diverse ways, including for task management and
personal archiving [2,13,23]. Foldering behaviors are the
most commonly studied email management practice.
Whittaker and Sidner [23] characterized three common
management strategies: no filers (forego using email
folders, relying on browsing and search), frequent filers
(minimize the number of messages in their email inbox by
frequently filing into many folders and relying on folders
for access), and spring cleaners (periodically clean their
inbox into many folders). Fisher et al. [11] also added a
fourth management strategy: users who kept their inboxes
trim by filing into a small set of folders. Other studies [1,2]
discovered similar management strategies, but also found
that users did not exclusively fall into one category. Rather,
users employ a combination of strategies over time [1,11].
Grouping messages together according to conversational
threads (i.e., a reply chain of messages on a common topic)
has been explored in prior research [2,3,19,20]. Gmail [12]
uses threads (rather than individual messages) as the basic
organizing unit for email management, although a more
recent version also combines the functionality of folders
and labels [16]. A thread-based inbox view is more
compact, enabling users to see more messages without
scrolling, helping those who rely on leaving messages in the
inbox to serve as ‘todo’ reminders. Collecting messages
into threads also gives users the context for interpreting an
individual message [19]. While Venolia and Neustaedter
Figure 1. User interface design for Bluemail, showing panes for foldering (A) and tagging (B), on the left, a message list area in the
top center showing a threaded message (C) and a selected thread (D) which is displayed in the message preview below showing an
interface to add tags to a message (E) and display tags already added to a message (F).
[19] and Bellotti et al. [2] conducted studies of threading
with small groups of users, there has not been a large-scale
study of thread usage.
One might think that the emergence of effective search
would lead users to reduce preparatory foldering. Yet
Teevan et al. [18] observed for web access that even a
perfect search engine could not fully satisfy usersneeds for
managing their information. Instead, their users employed a
mix of preparatory and opportunistic refinding behaviors.
We explore if this result holds for email refinding as well.
Other work has examined how people refind personal files
on their personal computers, showing that people are more
reliant on folder access than search. In addition, search and
navigation are used in different situations: search is only
used where users have forgotten where they stored a file,
otherwise they rely on folders [4]. Dumais et al. [9] found
that refinding emails was more prevalent than files or web
documents, and that refinding tended to focus on recent
emails. However, that study focused on search and did not
compare it to other access methods, e.g. folders or scrolling.
Elsweiler, et al. [10] looked at memory for email messages.
Participants were usually able to remember whether a
particular message was in their mailbox. Also, memory for
specific information about each message was generally
good; people remembered content, purpose, or task related
information best, correctly recalling over 80% of this type
of information, even when items were months old.
However, frequent filers tended to remember less about
their email messages. Filing information too quickly
sometimes led to the creation of archives containing
spurious information; premature filing also meant that users
were not exposed to the information frequently in the inbox,
making it hard to remember its properties or even its
Bluemail is the email client used for this study. It is a web-
based client that includes both traditional email
management features such as folders, and modern attributes
such as efficient search, tagging, and threads. This
combination of features allowed us to directly compare the
benefits of preparatory retrieval behaviors that rely on
folders/tags, with opportunistic search and threading. We
could not have made this direct comparison if we had used
a client such as Gmail that does not directly support folders
separately from tags. Also, Bluemail could be used to
access existing Lotus Notes emails, making the transition to
Bluemail very straightforward. For a full description of the
design see [17].
Figure 1 shows the main Bluemail interface. The layout
follows a common email pattern with navigation panes on
the left for views and foldering (to which Bluemail adds an
interface for tagging), a central content area with a message
list on top, and a message preview panel at the bottom.
Messages are filed into folders by drag and drop from the
message list into a folder in the left pane. One novel feature
of Bluemail that enhances scrolling is the Scroll Hint. As
the user engages in sustained scrolling (> 1 second) the
interface overlays currently visible messages with metadata
such as date/author of the message currently in view. This
hint provides orienteering information about visible
messages without interrupting scrolling.
Bluemail also supports efficient search (shown in the upper
right of Figure 1) based on a full content index of all
emails, with the search index being incrementally updated
as new messages arrive. As in standard email clients, and
unlike Gmail, messages can also be sorted by metadata
fields such as sender (‘who’), or date (‘when’). The default
view is by thread, which we now describe.
Message Threads
A message thread is defined as the set of messages that
result from the natural reply-to chain in email. In Bluemail,
threads are calculated against all the messages in a user’s
email database, i.e., threads include messages even if they
have been filed into different folders. This design contrasts
with clients that do not have true folders (like Gmail).
Bluemail uses the thread, not the individual message, as the
fundamental organizing unit. Deleting, foldering, or tagging
a thread acts on all the messages in the thread, even
messages already foldered out of view. Figure 1C shows
how threads are represented in the message list view. Each
thread is gathered and collapsed into a single entry in the
list. Users can toggle the view in the interface between the
default threaded view and the traditional flat list of
messages by clicking on the icon in the thread column
header. The ‘what’ column for a thread shows the subject
field corresponding to the most recently received message.
After the subject text, we show in gray text as much of the
message that space allows. User-applied tags are also
shown pre-pended to the subject in a smaller blue font, as
will be described in the tagging section below.
Tagging Messages
The interface for message tagging comprises four elements:
a tag entry and display panel in the message, pre-pended
tags in the list view’s what’ column, a tag cloud, and a
view of the message list filtered by tag. As a user tags
messages, the tags are aggregated into a tag cloud as shown
in Figure 1B. Clicking on a tag (anywhere a tag appears)
filters the message list to show only messages across a
user’s email (including other folders) with that tag. If any of
those messages are part of a thread, the whole thread is
shown in threaded view. Toggling to the unthreaded view
shows only the individual messages marked with the tag.
The Bluemail prototype was released in our organization
and used long term by many people. For our analyses, we
focused on frequent users, i.e., people who used our system
for at least a month, with an average of 64 days usage. As
our main focus was on access behaviors, a criterion for
inclusion was that a user had to have used each retrieval
feature (folder-access, scroll, search, sort, tag-access) at
least once. This assured us that users were aware of that
feature’s existence. Overall 345 people satisfied these
criteria. Users included people from many different job
roles (marketing, executives, assistants, sales, engineers,
communications) and organizational levels (managers and
non-managers). Unlike many prior email studies there were
few researchers (just 2% of our frequent users).
Many prior studies of email have taken a snapshot of a user
at a single point in time. This approach has the
disadvantage that it may capture the email system in an
atypical state. To prevent this, we therefore recorded
longitudinal daily system use, averaging measures across
the entire period that each person used the system.
General usage statistics
For each user, we collected and averaged the following
usage statistics over each day they used the system:
Days of system usage. We only included people with
more than 30 days of usage.
Total messages stored - number of messages included in
all folders and the inbox.
Inbox size - number of inbox messages.
Number of folders.
Messages per thread - number of messages in each
thread, excluding messages without replies.
Daily change in mailbox size. Other work notes that it is
hard to determine the exact numbers of received messages
because users delete messages [11]. We therefore
recorded the daily change in mailbox size, i.e., the
number of additional messages added or, in some cases,
removed from the total archive each day. From a
refinding perspective this is a better measure as it
represents the set of messages users potentially access
longer term.
Access behaviors
We also recorded various daily access behaviors. We
logged each instance when the behavior was invoked.
Sort - whenever the user clicked the various header fields
such as sender, subject, date, time, attachments, etc.
Folder-access - whenever a user opened a folder.
Scroll - whenever users scrolled for more than one second
(a conservative criterion adopted to identify when
scrolling is used for refinding).
Overall Usage Statistics.
Mean Std. Deviation
Days Used 63.97 42.61
Total Messages Stored 2568.79 3107.77
Inbox Size 870.28 1422.96
Number Folders 46.89 91.65
Messages/Thread 3.61 1.54
Daily Change in Size 24.24 58.07
Tag-access - whenever a user clicked on a tag.
Search - whenever the user conducted a search.
Open Message - whenever the user opened a message.
Operation duration - measured by subtracting the
timestamp of each operation from the timestamp of the
subsequent operation.
To preserve user privacy we did not record search terms or
the names of folders and tags. We initially recorded other
access operations, e.g., filter by flag (filtering for messages
users had marked as important), or filter by unread
messages (selecting the interface view which showed only
unread messages). However, these behaviors accounted for
less than 1% of all access behaviors and were only ever
used by 8% and 17% of our users respectively. We
therefore do not discuss them further.
We also recorded the success and duration of finding
sequences. We define a finding sequence as a set of access
behaviors containing one or more sort, scroll, search, tag-
access, or folder-access. Each finding operation was treated
separately, so that opening a folder followed by a sort was
treated as two separate operations. Searching followed by
sorting was treated the same way. Our analysis is
quantitative and relied on parsing large numbers of logfiles,
so we aimed to define an automatically implementable
definition of success and duration.
Success: People usually want to find a target message to
process the information it contains. We began by defining
as successful an unbroken sequence of finding operations
that terminated in a message being opened. Opening a
message did not always indicate success, however.
Observations of finding sequences revealed that users
sometimes opened a message briefly, discovered that it was
not the target, and then immediately resumed their finding
operations. To determine the upper bound for this
unsuccessful message opening interval, we timed 12 pilot
users opening and reading two standard paragraphs from an
email message that we felt would be sufficient for message
identification. We found this took 29s. Any ‘open message’
operation lasting less than 29s and followed by subsequent
finding operations was therefore treated as a non-terminal
part of the finding sequence. 23% of sequences contained
such unsuccessful opening of messages. Note that a user
briefly opening a message and hitting ‘reply’ would be
classified as a ‘success’ because the operation after ‘open
message’ is not a finding operation.
Failure: We classified as failures, sequences of finding
operations that did not terminate in a message being
opened, e.g., when the sequence was followed by the user
closing their browser, or composing a new message.
We acknowledge that finding success may also be
influenced by subjective factors such as urgency or message
importance. However our large-scale quantitative approach
requires clearly definable success criteria, and it is hard to
see how to operationalize these contextual factors in a
working logfile parser.
Duration: The finding sequence duration was the sum of
the finding operation durations it comprised. For one
specific case, we excluded final operation time: when
people abandoned an unsuccessful finding sequence, there
were sometimes long intervals, lasting tens of minutes
before the subsequent operation. We could not assume that
the user was actively engaged in that operation for the
entire interval, so we excluded it.
One potential limitation of this study is that we observed
behavior for people who have been using our system for an
average of two months. This may not be sufficient time for
people to modify long-term email behaviors. To
qualitatively profile our population however, we
interviewed 32 users. We found that 60% regularly used
Gmail, indicating that features such as tagging and search
were highly familiar. Furthermore, we ensured that all users
had used all access features at least once and found that
certain features such as threading were immediately used
ubiquitously—suggesting that people will readily change
access strategy if they see the value of new technology.
Overall Statistics
Table 1 shows overall usage statistics, derived from daily
samples. These are consistent with prior work (see
Whittaker et al. [21] for a review), showing that users tend
to build up large archives. However, the proportion (33%)
of messages we observed being kept in the inbox is smaller
than that reported in prior work. This may be due to
different sampling methods, i.e., that we were sampling
daily rather than relying on a single snapshot. Also, there
may be over-representation of researchers in prior samples,
and others [11] have speculated that researchers tend to
hoard more than other types of workers. Finally threads did
not tend to have a complex structure, with an average of
3.61 messages per thread, after we exclude singleton
messages (i.e., messages without replies). As with all prior
email research, there is high variability in most aspects of
usage, as shown by the large standard deviations.
Access Behaviors
We next examined people’s access behaviors, which have
not been systematically studied before.
. Daily Usage, Distributions and Durations for Each Access
Behavior. (Opportunistic behaviors are shaded.)
Mean (SD)
% of All
Mean (SD)
Duration in
0.21 (0.42) 12 58.82 (30.22)
Tag-accesses 0.02 (0.09) 1 not recorded
Searches 0.71 (1.82) 18 17.15 (36.86)
Sorts 0.17 (0.33) 7 13.96 (17.99)
Scrolls 1.49 (1.76) 62 25.77 (30.22)
Overall, each person had an average of 61.75 (SD 110.7)
finding sequences, lasting a mean of 69.59s (SD 33.48).
Overall 88% of finding sequences were successful, and the
average number of finding operations per sequence was
3.85 (SD 3.57). As expected, successful sequences
contained fewer operations than unsuccessful ones (M =
3.67, SD 1.83 and M = 5.03, SD 5.26), paired t-test t(344) =
9.41, p <0.001, presumably because on encountering failure
people persist in trying to find the target message.
Opportunistic behaviors dominate retrieval
Our first question concerned the overall frequencies of
different access behaviors. We can distinguish between: (a)
accesses based on preparatory activity, i.e., using folders
and tags that users deliberately create in anticipation of
retrieval, and (b) opportunistic accesses that do not rely on
preparatory activity, i.e. sorting, scrolling and searching.
Table 2 shows that opportunistic behaviors dominate. These
account for 87% of accesses. This is mainly explained by
the predominance of scrolling which accounts for 62% of
all accesses. Of course, scrolling might be used in
preparatory contexts, e.g. scrolling through a large folder.
However scrolling and folder-access are highly negatively
correlated (r(343)=-0.52, p<0.001). Overall then,
preparatory activities (folder- and tag-accesses combined)
are not prevalent. They account for just 13% of all access
operations overall.
A within sequence analysis indicated that specific behaviors
tended to perseverate, with people relying on one or two
strategies to find a specific message. Note too that there is
enormous variability in individual usage for each of these
behaviors (as indicated by their large standard deviations).
The use of tagging was minimal, accounting for just 1% of
all accesses. We therefore excluded it from subsequent
analyses, removing finding sequences that included tags
and relaxing the criterion that each user in the sample had
to have used tag-access at least once. This added 13 users to
our original user population.
Efficiency of Different Access Behaviors
Table 2 also indicates large differences in duration for the
different access behaviors. Each individual folder-access
took approximately one minute, more than twice as long as
scrolls, with both searches and sorts being relatively short
(around 15s). Paired t-tests show that folder-access
operations are significantly longer than scrolls (t(357) =
6.71, p<0.001). Scrolls in turn are significantly longer than
searches (t(357) = 2.87, p<0.01). However, there is no
difference between the durations of searches and sorts
(t(357) = 0.51, p>0.05).
Interrelations between Access Behaviors
We next explored the interrelations between access
behaviors, where we anticipated specific patterns. Users
who have made the effort to create folders should be more
reliant on a preparatory behavior like folder-access, and
avoid opportunistic behaviors like search, sort, and scroll.
Others we expected to rely exclusively on these
opportunistic behaviors, eschewing folder-access.
We found these expected combinations of access patterns.
When we correlated behaviors per user, preparatory
behavior, i.e. folder-access, was negatively correlated with
search (r(356) = -0.25, p<0.001), and with scrolling (r(356)
= -0.52, p<0.001). Thus, scrolling does not co-occur with
folder-access, e.g. scrolling through large folders. As we
expected, the various opportunistic behaviors positively
intercorrelate, with searches correlating with sorts: r(356) =
0.23, p<0.001), and with scrolls r(356) = 0.61, p<0.001).
This indicates that people tend to rely exclusively on either
preparatory or opportunistic behaviors, but not a mixture of
the two. This is an important result because it suggests that
email clients, that mainly support search, like Gmail, are
unlikely to be optimal for all users.
The Relationship between Email Management Strategy,
Threads and Access Behaviors
So far our analysis has only examined access behaviors. In
this section, we examine the relationship between access
behaviors, threads, and email management strategy. Are
people who engage in preparatory activity by making
folders more likely to rely on these for retrieval? We also
explore the effects of the intrinsic organization afforded by
threads. Do threads make people less likely to use folders
for access?
People who create folders use them more often for access
To analyze access behaviors with respect to management
strategy, we must first operationalize management strategy.
Prior work [1,13,23] proposed strategy differences, based
on a combination of inbox size, number of folders as well
as ‘large scale changes’ in inbox size over time. However,
recent work [11] critiques these definitions arguing that
they are ad hoc and do not reliably identify distinct user
types. To avoid both of these problems, we used a simple
propensity to organize metric based on the percentage of
the user’s total mailbox that is stored in folders. People who
are more committed to foldering should have a higher
percentage of their information in folders, as opposed to the
inbox. We conducted a median split on this percentage to
divide users into high and low filers.
Table 3 shows the frequency of different access behaviors
(folder-access, scroll, sort, search), depending on whether a
user is a high or low filer. To control for the large variance
Access behaviors for High and Low threading based on
median split of percent messages in threads.
% Mailbox
that is
% of Accesses
in which
Behavior is
SD Significance
High 10 18 Folder-
accesses Low 14 20
t(356) = 2.05,
p < 0.05
High 24 26 Searches
Low 13 18
t(356) = 4.73,
p < 0.001
High 7 10 Sorts
Low 7 9
t(356) = 0.09,
High 58 25 Scrolls
Low 66 26
t(356) = 2.65,
p < 0.01
Access behaviors for High and Low filers based on a
median split of percent messages in folders.
% Mailbox
% of Accesses
in which
Behavior is
SD Significance
High 16 22 Folder-
accesses Low 7 16
t(356) = 4.68,
p < 0.001
High 16 18 Searches
Low 21 26
t(356) = 2.05,
p < 0.05
High 8 10 Sorts
Low 6 9
t(356) = 1.95,
High 59 26 Scrolls
Low 65 25
t(356) = 2.39,
p < 0.02
in absolute numbers of accesses across different users, we
normalize by expressing each access behavior as a
proportion of all access operations for that user.
We expected that high filers would be more likely to use
folders at retrieval, and less likely to scroll, sort, or search.
This was confirmed: independent t-tests showed high filers
were more likely to use folder-access and less likely to
search or scroll (see Table 3). Contrary to our expectations,
filers were slightly more likely to sort, possibly after
accessing a large folder to identify a message from a
particular person, time or topic.
These are striking results because a median split is a
conservative statistical approach, as users who are just
above or below the median may be very similar in terms of
their filing strategy. We therefore checked our approach by
comparing upper and lower deciles (i.e., people who almost
always file with those who almost never file). To check the
validity of the normalization, we also compared absolute
numbers of each access behavior for the high/low split.
Both analyses replicated the basic findings reported above.
Intrinsic organization: Threading reduces reliance on folders
There are other factors, such as threading, which potentially
affect access behaviors. As shown in Figure 1, our client
automatically organizes and presents emails as threads.
Threading imposes a structure on messages and potentially
represents a way for people to access related messages,
without the burden of manually organizing them.
The perceived utility of threading is shown by the fact that
our users made almost exclusive use of the threaded view.
Users were able to switch from this view to a more standard
sequential message view, but seldom did so. For all system
users, 56% always used the threaded view, and for those
who switched to the unthreaded view, only 1% stayed there
more than 40% of the time.
We explored the effect of thread structure on access
behavior, using the same approach as for foldering strategy.
We first identified the percentage of each person’s
messages that participated in threads. We again conducted a
median split on this thread percentage to distinguish people
who received mostly heavily threaded emails (‘high
threading’) from those whose messages tended not to be
threaded (‘low threading’). We expected that those with
high degrees of intrinsic organization would be less reliant
on access methods that demanded manual preparation for
retrieval, such as folder-access.
Table 4 shows that, as we expected, people with highly
threaded emails are less likely to use folder-access. This
effect may be reinforced by the fact that in Bluemail,
threads include messages that have been foldered. People
who had more threads were also more reliant on search,
which is possibly a response to situations where threads
provide insufficient organization to access the message
people need. Finally, people with more threads were less
likely to scroll suggesting that threads were indeed an
effective way to compress information in the inbox.
Threads allow people to see more messages, reducing the
need to scroll.
How do Efficiency and Success Relate to Management
Strategy and Threads?
Foldering is Less Efficient and No More Successful
We next explored the overall efficiency of the preparatory
management strategy. It takes time and effort to manually
organize emails into folders, but does this effort pay off?
Do people who prepare find information more quickly and
successfully? Do they find information in fewer operations?
Table 5 reveals that high filers managed to find messages
using fewer operations in each finding sequence. However,
this did not equate to faster overall finding sequences, as
high filers took marginally longer in their finding
sequences. There is a simple explanation for this: high filers
are more reliant on folder-accesses, which Table 2 shows
take much longer than the searches and sorts.
More important is how often people successfully find the
target message. Again, to control for the fact that people
had very different numbers of finding sequences, we
evaluated what percentage of their finding sequences were
Success and efficiency of finding sequences for High and
Low threads based on median split of percent messages in threads.
Measure % Mailbox
that is
Mean SD Significance
High 0.91 .11 % of All
that are
Low 0.85 .12
t(356) = 2.01,
p < 0.05
High 66.43 27.20 Sequence
(secs) Low 72.35 37.44
t(356) = 1.71,
p > 0.05
High 4.03 2.55 #
Operations Low 3.81 1.44
t(356) = 0.98,
p > 0.05
Success and efficiency of finding sequences for High and
Low filers based on median split of percent messages in folders.
Measure % Mailbox
Mean SD Significance
High 0.88 .12 % of All
that are
Low 0.88 .11
t(356) = 0.98,
p > 0.05
High 72.87 38.05 Sequence
(secs) Low 66.07 26.64
t(356) = 1.97,
p < 0.05
High 3.69 1.46 # Operations
Low 4.16 2.50
t(356) = 2.17,
p < 0.05
successful. We expected high filers to be more successful
given their investment in preparing materials for retrieval
(‘I know where that message is because I deliberately filed
it’). As Table 5 shows, contrary to our expectations, high
filers were no more successful at finding messages than low
filers. Again we checked whether high vs. low filers had a
greater absolute success rate, but found no differences.
These analyses examine how management strategy affects
efficiency and success. However to confirm our analyses
we can also look directly at people’s access behaviors (as
opposed to their management strategy) to see how these
behaviors affect both efficiency and success. Thus for
efficiency, we would expect that people who were more
reliant on folder-access behaviors would tend to have
finding sequences that are longer in duration. To determine
which access behaviors dominated for each user, we again
calculated the frequency of each access behavior expressed
as a proportion of all accesses. We then correlated this with
the overall duration of their finding sequences. Consistent
with the above results, a high reliance on folder-access was
positively correlated with an increased sequence duration
(r(356)=0.33, p<0.001). Scrolls, sorts and searches were all
negatively correlated with sequence duration.
We next examined the relationship between retrieval
behavior and success. Did a reliance on folder-access
predict success, or was search a stronger predictor? We
found that people who relied on search were more likely to
have successful finding sequences (r(356)=0.15, p<0.005).
None of the other behaviors was correlated with success.
Threads improve finding success
We also explored whether the intrinsic organization
afforded by threads improved success and efficiency. Table
6 shows that people who had higher threading were more
successful at finding messages. Threads did not seem to
influence efficiency: there were no differences in either
sequence duration or operations per sequence.
Explaining Management Strategy
Given that actively creating folders does not increase
efficiency or success, why do some people use them? One
possibility is that they are a task management strategy.
Many people use the inbox as a ‘todo’ list, a function which
is compromised by a high incoming message volume,
causing them to folder. Foldering removes messages from
the inbox, reduces its complexity, and allows users to see
outstanding tasks at a glance. To explore whether foldering
is used for task management, we compared incoming
message volume to users’ propensity to folder. Our measure
of incoming volume was the daily change in inbox size, i.e.,
how many messages people kept each day. We correlated
this with our standard measure of people’s propensity to
folder, i.e., proportion of the mailbox in folders. We found
that people who kept more messages each day were more
likely to folder (r(356)=0.16, p< 0.01). This suggests that
foldering may be a reaction to incoming message volume.
To further understand this result, we asked our 32 interview
participants about their email management and refinding
practices. Though people used their folders to find
messages, the predominant reason given for foldering in the
first place was related to task management, as comments
from four different folder users illustrate: “Generally
everything sits in inbox until actioned… [I] attempt a daily
run through to move inbox items to subfolders.” “My inbox
is a todo list.” “I’m trying not to drown.” “My inbox stays
clean. It has things I need to respond to or do.”
A related possibility is that assiduous filing is the result of a
greater need to refind messages due to a user’s job role,
regardless of whether filing is less efficient. However our
data do not support this hypothesis: there were no overall
differences in the number of accesses conducted by filers
and no filers (t(356)=1.05, p > 0.05). To confirm this we
also directly explored whether job role affects strategy, but
found few effects. Although managers get more messages
(t=3.77, p<0.001), when compared with non-managers they
show no overall differences in refinding behaviors (sorts
(t=1.49), searches (t=.14), folder-access (t=.80), scrolls
(t=.08)) or in their propensity to folder (t=.25) (all df=356).
This is the first large-scale quantitative study of how people
refind email. Our results show that opportunistic access
behaviors dominate, mainly because of the prevalence of
scrolling. Nevertheless, folder-accesses account for 12% of
overall accesses. Thus even with a client that supports
effective search and threading, some users are still reliant
on folder-accesses.
Prior work has argued that folders may be poorly organized
and sometimes ill-suited for retrieval [23]. While people
who manually organize more information into folders are
more likely to rely on these for retrieval, high filers were no
more successful at retrieval. Further, they were less
efficient because folder-accesses took longer on average.
Why then, do users persist with manual foldering when it is
known to be onerous to enact [3]? First, it was clear that
foldering was not a response to increased demands for
refinding emails; filers were no more likely to reaccess
messages. Instead we found that filing seems to be a
reaction to receiving many messages. Users receiving many
messages were more likely to create folders, possibly
because this serves to rationalize their inbox, allowing them
to better see their ‘todos’. Interview data confirms that
people file to clean their inboxes to facilitate task
management. This result contradicts prior work arguing that
people who receive many messages do not have the time to
create folders [1]. Further work should move beyond
refinding to explore trade-offs between opportunistic and
preparatory strategies for task management.
We also found that the intrinsic structure afforded by
threads affected folder-access. People who received more
threaded messages were less likely to rely on folders for
access. Threads impose order on the mailbox, reducing the
need for preparatory strategies. In part, this validates our
design. Threading in Bluemail draws messages out of
folders and into relevant inbox threads, making people less
reliant on folders for access. Threads also serve to compress
the inbox, reducing the amount that users need to scroll. As
a result, people who received more threaded emails were
more successful in their retrievals.
There are direct technical implications of our results.
Search was both efficient and led to more successful
retrieval, in part supporting the search-based approach of
clients like Gmail. However in our study, other behaviors,
especially scrolling, were prevalent. Gmail, which mainly
supports search at the expense of scrolling, foldering, and
sorting may be suboptimal. Even with a threaded client,
scrolling was by far the most common access mechanism.
However, scrolling is not well supported in Gmail, which
breaks the mailbox into multiple pages, each of which has
to be accessed and viewed separately. Gmail also does not
support sorting, although this was a less frequent access
behavior. Finally, folder-access was a preference for a
minority of users, accounting for 12% of accesses
(compared with 18% that were searches). Recent versions
of Gmail attempt to combine folders and search [16].
However our data argue for the opposite: users employ
either preparatory or opportunistic approaches, suggesting
we need to design different features or mailbox views that
optimize for each population tendency.
Another important design implication concerns threading
which proved very useful. People who received more
densely threaded emails created fewer folders and relied
less on folder-accesses. They were also more successful at
accessing emails. Threading can be improved, however. At
the moment, threading imposes a very low level of
organization [21]. The average thread length we observed
was just 3.61 messages, and only 16% of messages
participate in threads. This suggests there may be room for
different intrinsic organizational tools that collate larger
numbers of messages around a task.
How might we impose higher-level intrinsic organization
on email? One possibility is to re-organize the inbox
according to ‘semantic topics’. One could use clustering
techniques from machine learning to organize the inbox
into ‘superthreads’ by combining multiple threads with
overlapping topics, using techniques similar to [8].
In addition to ‘superthreads’, there may be other
opportunities to exploit intrinsic organization to reduce the
burden of manual organization. Several systems organize
emails on the basis of social information, such as key
contacts and social networks [14,22]. This approach has led
to new commercial clients that include these features, such
as Xobni and newer versions of Outlook. However, we
currently lack systematic data about the utility of these new
socially organized clients.
Finally, there are important empirical and theoretical links
to other areas of PIM. Our findings that people resist using
tags to manage emails are consistent with PIM studies
showing people are unwilling to use tags for organizing
personal files [6]. However recent studies also demonstrate
that powerful new search features do not cause people to
abandon manual navigation to desktop files [4] or web
documents [18]. In contrast, we found a preference for
newer automatic methods, such as search and threading,
and that these were more effective than manual techniques.
This may be because email data is more structured than
personal files and webpages, leading to more effective
searches. Another possibility is that the volume of email
messages received is high compared with files created or
web pages visited, making manual organization too onerous
for email. Future work needs to explore this more.
In conclusion, we have presented a study that contributes a
deeper understanding of email message refinding, a topic
that has not yet been systematically studied. We have
extended prior studies that focused on snapshots of email
data. We have provided new data about the relations
between management strategy, intrinsic threading structure,
and actual access behaviors. We have also shown the value
of threading and search tools. These data also offer direct
design implications for current and future clients, including
improved scrolling and threading. In future work, we will
explore new forms of intrinsic organization that our results
here suggest.
1. Bälter O. Keystroke Level Analysis of Email Message
Organization. Proc. of CHI 2000, 105-112.
2. Bellotti, V., Ducheneaut, N., Howard, M., & Smith, I.
Taking email to task: The design and evaluation of a
task management centered email tool. Proc. of CHI
2003, 345-352.
3. Bellotti, V., Ducheneaut, N., Howard, M., Smith, I., &
Grinter, R. Quality vs. quantity: Email-centric task-
management and its relationship with overload.
Human-Computer Interaction, 20(1-2), 2005, 89-138.
4. Bergman, O., Beyth-Marom, R., Nachmias, R.,
Gradovitch, N., & Whittaker, S. Advanced search
engines and navigation preference in personal
information management. TOIS, 26(4), 2008, 1-24.
5. Boardman, R., Sasse, M.A. ‘Stuff goes into the
computer and doesn’t come out’: A cross-tool study of
personal information management. Proc. of CHI 2004,
6. Civan, A., Jones, W., Klasnja, P., Bruce, H. Better to
organize personal information by folders or by tags?:
The devil is in the details. Proc. of ASIS&T, 45(1)
2008, 1-13.
7. Dabbish, L.A., Kraut, R.E., Fussell, S., Kiesler, S.
Understanding email use: Predicting action on a
message. Proc. of CHI 2005, 691-700.
8. Dredze, M., Lau, T., Kushmerick, N. Automatically
classifying emails into activities. Proc. of IUI 2006, 70-
9. Dumais, S., Cutrell, E., Cadiz, J., Jancke, G., Sarin, R.,
Robbins, D. Stuff I’ve seen: A system for personal
information retrieval and re-use. Proc. of SIGIR 2003,
10. Elsweiler, D., Baillie, M., Ruthven, I. Exploring
memory in email refinding. TOIS, 26(4), 2008, 1-36.
11. Fisher, D., Brush, A.J., Gleave, E., Smith, M.A.
Revisiting Whittaker & Sidner’s ‘email overload’ ten
years later. Proc. of CSCW 2006, 309-312.
12. Gmail, About Gmail.
help/about.html, verified Jan. 11, 2011.
13. Mackay, W., Diversity in the use of electronic mail: A
preliminary inquiry. TOIS, 6(4), 1988, 380-397.
14. Neustaedter, C., Brush, A.J., Smith, M., Fisher, D. The
social network and relationship finder: Social sorting
for email triage. CEAS, 2005.
15. Radicati Group, Email statistics report 2010-2014., verified Jan. 11,
16. Rodden K., Leggett, M. Best of both worlds:
Improving Gmail labels with the affordances of
folders. Extended Abstracts of CHI 2010, 4587-4596.
17. Tang, J.C., Wilcox, E., Cerruti, J.A., Badenes, H.,
Schoudt, J., Nusser, S. Tag-it, snag-it, or bag-it:
Combining tags, threads, and folders in e-mail,
Extended Abstracts of CHI 2008, 2179-2194.
18. Teevan, J., Alvarado, C., Ackerman, M.S., Karger,
D.R. The perfect search engine is not enough: A study
of orienteering behavior in directed search. Proc. of
CHI 2004, 415-422.
19. Venolia, G., Neustaedter, C. Understanding sequence
and reply relationships within email conversations: A
mixed-model visualization. Proc. of CHI 2003, 361-
20. Wattenberg, M., Rohall, S.L., Gruen, D., Kerr, B.
Email Research: Targeting the enterprise. Human-
Computer Interaction, 20, 1 & 2, 2005, 139-162.
21. Whittaker, S., Bellotti, V., & Gwizdka, J. Everything
through email. In W. Jones, J. Teevan (Eds.), Personal
information management. UW Press, 2007, 167-189.
22. Whittaker, S., Jones, Q., Nardi, B., Creech, M.,
Terveen, L., Isaacs, E., Hainsworth, J. ContactMap:
Organizing communication in a social desktop. TOCHI
11, 4, 2004, 445-471.
23. Whittaker, S., Sidner, C. Email overload: Exploring
personal information management of email, Proc. of
CHI 1996, 276-283.
... Bluemail [102], [17], Squadbox [75], Themail [21] Streaming Platforms (e.g., Twitch, YouTube Live, DouYu) BabyBot [91], Snapstream [105], StreamSketch [70], StreamWiki [69], VisPoll [14] Slack ...
... For some system designs, the novel components being evaluated primarily impact only one person, allowing for others using the social system to be unaware of the study. For example, the Bluemail email client contributed new approaches for email searching and categorizing, requiring little need for engagement with the people with whom participants were sending or receiving emails [102]. Similarly, when FeedMe introduced a new technique for directed content sharing, it allowed recipients to receive shared content over emails they already used rather than sign up for the service themselves [7]. ...
Full-text available
The CSCW community has a history of designing, implementing, and evaluating novel social interactions in technology, but the process requires significant technical effort for uncertain value. We discuss the opportunities and applications of "piggyback prototyping", building and evaluating new ideas for social computing on top of existing ones, expanding on its potential to contribute design recommendations. Drawing on about 50 papers which use the method, we critically examine the intellectual and technical benefits it provides, such as ecological validity and leveraging well-tested features, as well as research-product and ethical tensions it imposes, such as limits to customization and violation of participant privacy. We discuss considerations for future researchers deciding whether to use piggyback prototyping and point to new research agendas which can reduce the burden of implementing the method.
... PIM is user behavior in which people collect information in order to refind and use them later on [3,4]. PIM research studies the collection and retrieval of emails e.g., [5,6], file documents [7,8], task management [9,10], personal health records [11,12], personal photo collections [13,14], and website refinding [15,16]. The two last PIM domains share similar attributes with music collections, which they do not share with other PIM domains. ...
Full-text available
Streaming music application users can collect music (e.g., by “liking” songs or adding them to playlists) at no additional cost. However, most people reduced their collection size drastically when moving to streaming technology. The aim of the current study was to determine whether increasing streaming collection size would cause an increase in listening enjoyment. We conducted a within-subject controlled experiment where we asked our 40 participants, who have small collections, to rate their listening enjoyment in three different conditions: baseline condition — regular listening without manipulation, experimental condition — where they were asked to collect music and listen to it, and platform condition — where they were asked to listen to the songs suggested to them by their application algorithm. Listening enjoyment was rated in real time to avoid memory bias. Results indicate that (a) the participants’ current listening enjoyment was lower than they estimated it was, and seems in need of improvement; (b) collecting songs was more pleasurable than difficult; and (c) listening to collected songs significantly increases listening enjoyment and reduces enjoyment dropdowns over both baseline and platform condition. These results strongly suggest that streaming music providers should encourage their users to collect music. In our discussion, we shed light on a fundamental design problem that most streaming applications have which might discourage music collecting, and suggest designs aimed to encourage collecting.
... For example, recurrent themes of discussion include the use of folders versus tags (Bergman et al., 2013a;Civan et al., 2008;Voit et al., 2012) and navigation versus search (Bergman et al., 2008(Bergman et al., , 2013bFitchett and Cockburn, 2015;Teevan et al., 2004). Many excellent studies focus on how people use and organize specific forms of information, e.g., their email (Bellotti et al., 2003;Capra et al., 2013;Hanrahan and Pérez-Quiñones, 2015;Whittaker et al., 2011) and bookmarks (Abrams et al., 1998;Boardman and Sasse, 2004;Jones et al., 2002), but this also makes the field of PIM fragmented. PIM is also related to the notion of quantified self, which is concerned with the tracking of personal activities, often through a dedicated hardware device (e.g., physical fitness monitors and activity trackers such as smartwatches) (Gurrin et al., 2014). ...
This paper presents an ecosystem for personal knowledge graphs (PKG), commonly defined as resources of structured information about entities related to an individual, their attributes, and the relations between them. PKGs are a key enabler of secure and sophisticated personal data management and personalized services. However, there are challenges that need to be addressed before PKGs can achieve widespread adoption. One of the fundamental challenges is the very definition of what constitutes a PKG, as there are multiple interpretations of the term. We propose our own definition of a PKG, emphasizing the aspects of (1) data ownership by a single individual and (2) the delivery of personalized services as the primary purpose. We further argue that a holistic view of PKGs is needed to unlock their full potential, and propose a unified framework for PKGs, where the PKG is a part of a larger ecosystem with clear interfaces towards data services and data sources. A comprehensive survey and synthesis of existing work is conducted, with a mapping of the surveyed work into the proposed unified ecosystem. Finally, we identify open challenges and research opportunities for the ecosystem as a whole, as well as for the specific aspects of PKGs, which include population, representation and management, and utilization.
... Interestingly, very few participants had strategies to manage their university email account and indeed there is research to suggest that complex strategies may offer little over and above a simple search and retrieve approach (Whittaker et al. 2011). Students felt that email communication was mass communication and contained generic rather than personally relevant information and as such rarely necessitated a response. ...
Technology to enable and support learning and teaching is widespread in university settings. One consequence of such technology use is the accumulation of large volumes of digital data. The acquisition of, and failure to, discard digital content can lead to digital clutter. The potential negative consequences of digital clutter have been examined mainly within a workplace context. Far less is known about how university students manage their academic digital data and whether they have strategies to deal with excessive digital clutter. Eighteen undergraduate students took part in a one-to-one or group-based interview to discuss their digital data management strategies including accumulation and deletion behaviours. Thematic analysis led to three themes: (1) Digital data accumulation across the student journey, (2) Reactive and evolving digital data management strategies and (3) Data overload: Anxiety, loss of productivity and feeling overwhelmed. The findings capture the complexity of feelings students have about different types of digital technology and the strategies they use to manage increasingly large volumes of digital data. Findings are discussed in relation to the need for better support and guidance for students around the use of digital technology to manage their data during their time at university.
... 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]). ...
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.
... Discourse shows that one of the most common tools for effectively managing personal information management is e-mail (Whittaker et al. 2006;Whittaker, et al., 2011;Whittaker and Sidner, 1996). One component that may be particularly pertinent to the aim of understanding personal records management at home is the discussion and research regarding the use of e-mail as a de-facto personal information management tool (Ducheneaut and Bellotti, 2001, p. 37). ...
Full-text available
Introduction This paper considers how we can better manage personal records in the home b addressing questions such as how and why personal records are retained in an electronic form and how they are managed Method A qualitative method with semi-structured interviews was used. Participants were recruited through social media. The interviews included virtual guided tours of personal records.There were thirty participants in twenty-two interviews (some interviews were with couples). Analysis Each stage of the personal records management process described by participants was observed and categorised, resulting in an inclusive flow diagram. Results The management of personal records at home can be categorised and described in terms of a flow. Some commonalities were found between personal information management in the workplace and at home, such as the frequent use of e-mail to manage records and the use of micro-notes and reminders. Conclusion Personal records management at home can be described as a flow through which records progress. The fact that the study of personal information management has rarely addressed personal information management at home offers many opportunities for fruitful future research.
... Our current study indicates that collection size and retrieval tools influence the strategies that people use to organize and retrieve photos compared with those deployed for personal files. Future work could explore how collection size and tools influence organization and retrieval of other types of PIM items such as emails [61] or web pages [16]. ...
Full-text available
We tested the use of smartphones for retrieval of pictures of long-term, salient family events. Our goal was to replicate a study conducted a decade ago where participants used digital cameras. We found that smartphones affected picture retrieval in two contrasting ways. Overall, the constant availability of smartphones increased collection size. This increased the failure percentage for pictures downloaded to computer file system folders from an average of 43% in the original study to 71% in the current one. On the other hand, improved smartphone retrieval technologies including timeline, search, and face recognition reduced smartphone application retrieval failures to 29% on average. Overall, these two opposing tendencies canceled each other out, with no significant difference in failure percentage and retrieval time between the two studies. Results indicate that the magnitude of pictures is too much for us to manually handle and we must rely on technology for picture retrieval.
The CSCW community has a history of designing, implementing, and evaluating novel social interactions in technology, but the process requires significant technical effort for uncertain value. We discuss the opportunities and applications of "piggyback prototyping", building and evaluating new ideas for social computing on top of existing ones, expanding on its potential to contribute design recommendations. Drawing on about 50 papers which use the method, we critically examine the intellectual and technical benefits it provides, such as ecological validity and leveraging well-tested features, as well as research-product and ethical tensions it imposes, such as limits to customization and violation of participant privacy. We discuss considerations for future researchers deciding whether to use piggyback prototyping and point to new research agendas which can reduce the burden of implementing the method.
Full-text available
The purpose of this article is to identify the key attributes of personal electronic records in order to develop systems that may enable people to manage them in the home. As more personal information becomes electronic, this is increasingly necessary. Personal electronic records were identified and categorised using interviews and virtual guided tours. Three main attributes were identified: primary user-subjective categories; attributes which identify the circumstances that give rise to the records; and attributes which describe the legal validity of each record. In addition to providing an improved understanding of personal electronic records in the home, these attributes are developed into a set of potential metadata fields.
Full-text available
The information society was defined by Daniel Bell (1976) as a post-industrial society where information leads processes of change and innovation. The constant flow of information in the 21st century is proving him correct. Today, people confront myriad streams of information from an overwhelming number of information channels. In the last decade, the amount of digital information has doubled every year (Jacobs, 2013; Schull, 2018). The information explosion, driven by the rapid development of new information technologies, has changed interactions between people and information and between people and the management of their own personal information (Bergman & Whittaker, 2016; Jones & Teevan, 2007). Personal information management (PIM) refers to the practices people use to acquire, organize, maintain, and retrieve information items as part of their daily routine (Jones, 2007). The personal experience of people with the vast amount of information available today manifests in changes in PIM practices (Jones et al., 2015), and in the emotions and feelings characterizing this encounter (Cushing, 2012). Most studies have focused on the use of specific PIM practices (e.g., filing and piling, navigating between folders, searching), ignoring the equally important affective aspects. Little work has examined the gaps between actual PIM behavior (how people manage information) and ideal PIM behavior (how people would like to manage information). Nor do previous studies offer a theoretical framework capturing a range of affective interactions between people and information or defining the whole affective experience of PIM. Study’s goals The study aimed to examine the affective aspects of people’s interactions with personal information and its management, with a sample of male and female participants of various ages. The research had four main goals. The first was to define the characteristics of the gaps between actual and ideal PIM behavior, drawing on Higgins’ self-discrepancy theory (1987). The second was to characterize the affective aspects of PIM and define their frequencies. The third was to examine the relations between the affective aspects of PIM and the use of practices (actual practice, ideal practices, and the gaps between them). The last goal was to define a typology of PIM behaviors and examine the affective aspects relating to them. Research questions A conceptual framework was constructed and validated to examine the affective aspects of PIM. This included the representation of the variables and their relations in two circles. The inner circle described sets of practices for the management of personal information (actual practice, ideal practices, gaps between them). The outer circle comprised the seven affective aspects of PIM: anxiety, efficacy, frustration, desperation, belonging, dependence and loss of control. It included three independent variables characterizing people who manage personal information: use of PIM platforms, age, and gender. Four research questions arose from the conceptual framework: (1) What is the actual and ideal use of PIM practices and what are the gaps between them? (2) What are the affective aspects of PIM? (3) What are the relations between the affective aspects and PIM practice use (actual, ideal, gaps)? (4) What are the types of PIM behaviors and their characteristics? Methodology A mixed methods approach was used to examine the affective aspects of PIM. This approach enables researchers to draw on the strengths of both qualitative and quantitative methods to deepen the observation of the research questions (Creswell, 2015). Most data were quantitative; the qualitative data were used to support the findings and suggest their meaning (Creswell, 2009). Quantitative data were used for the following proposes: measure actual and ideal practices and the gaps between them; examine the affective aspects of PIM; to define the relations between affective aspects and practices; define a typology of PIM behaviors. Qualitative data supported the findings by describing the affective experience of participants and giving examples of PIM behavior types. Participants included 465 respondents, 351 female and 114 male, aged 19-73. They filled in two questionnaires developed and validated for the study: a PIM practice questionnaire examining actual and ideal use of 25 practices, and a PIM affective experience questionnaire addressing the seven identified affective aspects. In addition, 16 in-depth one-hour interviews with eight females and eight males were recorded and transcribed. Quantitative data were analyzed via IBM-SPSS statistical software. Descriptive and explanatory procedures included Pearson correlation, t-tests for related and independent samples, two-step cluster analysis, and ANOVA. Qualitative data were analyzed using the Moustakas (1994) content analysis method, including horizontal and cluster of meanings analysis. Findings Findings revealed that the PIM experience becomes complex for people who manage personal information spaces in a digital, overloaded, and connected world. Participants were unsatisfied with their PIM behavior and wished to conduct more PIM practices that would enable them to reduce clutter and overload in the personal information space. Gaps between actual and ideal behavior were expressed in a complex and intense affective experience characterized by anxiety and frustration on the one hand but by a sense of high efficacy and little desperation on the other. The first research question examined the gaps between actual and ideal PIM behavior. Findings showed significant gaps between actual and ideal use for most practices. These gaps were mostly positive, revealing that participants wished to use more practices than they actually did and were not satisfied with their PIM behavior. Gaps were extremely large for organizing practices. Women had larger gaps between actual and ideal behavior; these were related to negative feelings and decreased with age. The second research question explored the affective experience of PIM. Findings revealed participants had high levels of anxiety when thinking about a possible loss of personal information or a failure of their digital platforms. They felt dependent on their personal information, were concerned about the amount of personal information they accumulated, and questioned their ability to organize it. On a more positive note, they expressed a sense of high efficacy in managing their personal information, seldom felt desperation, and felt more in control of their personal information than expected. Similar to the gender and age differences in the PIM practices, the affective experience was more intense for female participants and decreased with age. The third research question examined the relations between the affective aspects of PIM and the PIM practice use. Findings showed the affective experience of PIM was correlated mainly with PIM ideal behavior. For example, the more the participants felt anxiety, frustration, belonging, and dependency, the less they deleted information items and the more they filed emails in folders. With increased anxiety and frustration, participants wanted to use PIM practices less. The fourth research question examined types of PIM behaviors. The cluster analysis indicated four types of PIM behaviors differing by activity level (actual PIM) and satisfaction level (ideal PIM and gaps): passive and satisfied, active and satisfied, fairly active and unsatisfied, and active and fairly satisfied. Types differed in their range of affective aspects and in their use of digital PIM platforms. Conclusions and implications The study showed that people do not give up on managing their personal information spaces, despite the growing challenges posed by the information explosion and the divergent, multiple information technologies. However, it is impossible to ignore the intense negative affective experience accompanying PIM, including feelings of anxiety and frustration, or the substantial gaps between actual and ideal PIM behavior. Gaps can motivate people to actively reduce the discrepancies between actual and ideal behavior, but large gaps could harm motivation to manage the personal information space and trigger a more intense experience, especially in the unsatisfied type of information management behavior. The study has implications for research, consumer training, and platform design. Theoretically, it suggests self-discrepancy theory and the theoretical framework of the affective aspects of PIM could be useful in future PIM study and HCI research. In a more practical sense, it suggests principles for a training program to improve people’s PIM literacy skills. Finally, it indicates the need for platform designers to develop affective-sensitive and type-sensitive digital platforms for PIM.
Full-text available
ABSTRACT It is widely acknowledged,that many,professionals suffer from “e-mail overload.” This article presents findings from in-depth fieldwork that examined,this phenome- non, uncovering six key challenges of task management in e-mail. Analysis of quali- tative and quantitative data suggests that it is not simply the quantity but also the col- HUMAN-COMPUTER INTERACTION, 2005, Volume 20, pp. 89–138 Copyright © 2005, Lawrence Erlbaum Associates, Inc. Victoria Bellotti is a social scientist with an interest in computer-mediated com-
Full-text available
The research program at IBM's® Collaborative User Experience (CUE) group supports an e-mail system used by millions of people. We present three lessons learned from working with real-world enterprise e-mail solutions. First, a pragmatic, system-level approach reveals that e-mail programs are generally used idiosyncratically, often for many different goals at once. This fact has strong implications for both the design and assessment of new features. Second, we discuss how viewing e-mail as an element of corporate collaboration--not just communication--provides insights into problems with current systems as well as potential solutions. Third, we describe constraints imposed by the realities of software development and how they shape the space of feasible new designs. Finally, we illustrate these lessons with an overview of CUE research strategies in the context of an extended case study of one specific new technology: Thread Arcs. Although not all researchers work with an enterprise-level product team, we believe the experiences described here will be useful to anyone wishing to see their ideas ultimately implemented on a broad scale.
Conference Paper
Full-text available
Email triage is the process of going through unhandled email and deciding what to do with it. This process can quickly become a serious problem for users with large volumes of email. Studies have found that people use a variety of approaches to triage their email, many of which have a social component. We believe that email clients can better support email triage by providing users with additional sorting features based on socially salient information. We present a prototype email client, SNARF (the Social Network and Relationship Finder), that aggregates social meta-data about email correspondents to aid email triage. Users can then sort their correspondents based on this meta-data, bringing emails from socially important people to the foreground.
Conference Paper
Full-text available
This paper reports a study of Personal Information Management (PIM), which advances research in two ways: (1) rather than focusing on one tool, we collected cross-tool data relating to file, email and web bookmark usage for each participant, and (2) we collected longitudinal data for a subset of the participants. We found that individuals employ a rich variety of strategies both within and across PIM tools, and we present new strategy classifications that reflect this behaviour. We discuss synergies and differences between tools that may be useful in guiding the design of tool integration. Our longitudinal data provides insight into how PIM behaviour evolves over time, and suggests how the supporting nature of PIM discourages reflection by users on their strategies. We discuss how the promotion of some reflection by tools and organizations may benefit users.
Conference Paper
Full-text available
Email is one oftl~ most successful computer applicmiom yet devised. Our empin~:al ct~ta show however, that althongh email was origiraUy designed as a c~nmunica/ons application, it is now used for ~tional funaions, that it was not designed for, such as tab management and persona/ afoOt/v/rig. We call this ernt~l oveHoad We demonstrate that email overload creates problems for personal information manageaa,cnt: users eden have cluttered inboxes cor~mining hundreds of n~:age~¢, incl~rling outstanding tasks, partially read documents and conversational threads. Furthermore,, user attemt:Xs to rationalise their inbox~ by ~ing are ~Ron unsuccessful, with the consequence that important rr~ges get overlooked, or "lost" in archives. We explain how em~l over/oad/ng arises and propose technical solutions to the problem.
Conference Paper
Full-text available
Email has come to play a central role in task management, yet email tool features have remained relatively static in recent years, lagging behind users? evolving practices. The Taskmaster system narrows this gap by recasting email as task management and embedding task-centric resources directly in the client. In this paper, we describe the field research that inspired Taskmaster and the principles behind its design. We then describe how user studies conducted with ?live? email data over a two-week period revealed the value of a task-centric approach to email system design and its potential benefits for overloaded users.
Modern work is a highly social process, offering many cues for people to organize communication and access information. Shared physical workplaces provide natural support for tasks such as (a) social reminding about communication commitments and keeping track of collaborators and friends, and (b) social data mining of local expertise for advice and information. However, many people now collaborate remotely using tools such as email and voicemail. Our field studies show that these tools do not provide the social cues needed for group work processes. In part, this is because the tools are organized around messages, rather than people. In response to this problem, we created ContactMap, a system that makes people the primary unit of interaction. ContactMap provides a structured social desktop representation of users' important contacts that directly supports social reminding and social data mining. We conducted an empirical evaluation of ContactMap, comparing it with traditional email systems, on tasks suggested by our fieldwork. Users performed better with ContactMap and preferred ContactMap for the majority of these tasks. We discuss future enhancements of our system and the implications of these results for future communication interfaces and for theories of mediated communication.
People can organize information items by placing them into folders or by tagging them with labels. Over the years there has been considerable discussion on the relative merits of folders vs. labels. But there is relatively little empirical data directly comparing people's experiences with each model of organization on comparable sets of information. We compared participants' experiences organizing information with both folders and labels over time. Results indicate that each model has its strengths and weaknesses when assessed with respect to basic activities of personal information management (PIM) such as keeping, organizing and re-finding. Results afford a deeper, more detailed understanding of each model in practice. This understanding is achieved by engaging participants in an active, “hands-on” comparison of models over a period of time. Study results point to syntheses in tool support for organizing information that might combine strengths of folders and labels while avoiding the inadvertent introduction of weaknesses from these models.
Conference Paper
Organization of email messages takes an increasing amount of time for many email users. Research has demonstrated that users develop very different strategies to handle this organization. In this paper, the relationship between the different organization strategies and the time necessary to use a certain strategy is illustrated by a mathematical model based on keystroke-level analysis. The model estimates time usage for archiving and retrieving email messages for individual users. Besides explaining why users develop different strategies to organize email messages, the model can also be used to advise users individually when to start using folders, clean messages, learn the search functionality, and using filters to store messages. Similar models could assist evaluation of different interface designs where the number of items increase with time.