Methods&Tools p27-39

Methods&Tools p27-39 6/7/99 6:37 AM Page 27
V. KAPTELININ, B. NARDI AND C. MACAULAY methods & tools
David Wasserman ©1997 Artville, LLC
The Activity Checklist: A Tool for
Representing the “Space” of Context
Introduction
IIn recent years, specialists in human–computer interaction (HCI) have come to appreciate
the importance of understanding the context in which computer-supported activities take
place [1]. Such understanding directly affects design and evaluation by revealing what
users are up to and how they might most effectively use a technology. The idea is to gain
this understanding before the design process has progressed too far, or during evaluation,
when openings for modifications and improvements to the technology exist .
i n t e r a c t i o n s . . . j u l y + a u g u s t 1 9 9 9
27

Methods&Tools p27-39 6/7/99 6:37 AM Page 28
There have been several attempts to come
tual system of activity theory for the specific
up with tools and techniques to support
tasks of design and evaluation.
taking context into account in the design and
The Activity Checklist is intended to elu-
evaluation of computer technologies. These
cidate the most important contextual factors
approaches include task analysis [6], participa-
of human–computer interaction. It is a guide
tory design [3], and contextual design [7],
to the specific areas to which a researcher or
Victor Kaptelinin
among others. However, contextual factors are
practitioner should be paying attention when
Department of
notoriously elusive and difficult to pin down
trying to understand the context in which a
Informatics
[5], so there is still a need for conceptual tools
tool will be or is used. The Checklist lays out
Umeå University
to deal with context at a practical level.
a kind of “contextual design space” by repre-
S-901 87 Umeå, Sweden
The existing approaches to context are for
senting the key areas of context specified by
Victor.Kaptelinin@
the most part “bottom up” ones. They start
activity theory.
informatik.umu.se
with an empirical analysis of contextual fac-
In the rest of this paper we discuss activity
tors and gradually develop concepts such as
theory, present the Checklist, and show its use
Bonnie A. Nardi
“task decomposition” [6], “future workshops”
by giving an example of a specific technology.
Department of
[3], or “flow models” [7], which later can be
The Checklist is an adjunct to the basic prin-
Human–Computer
put in an appropriate theoretical framework.
ciples of activity theory—not a tool to be used
Interaction
From our point of view, this “bottom up” or
in isolation. An overview of activity theory
AT&T Labs–Research
empirically-driven strategy can be comple-
with empirical applications can be found in
Menlo Park, California
mented with a “top down” one, that is,
[13].
94040, USA
starting with an abstract theoretical represen-
nardi@research.att.com
tation of context and then situating this repre-
Basic Principles of Activity Theory:
sentation in the reality of design and
An Overview
Catriona Macaulay
evaluation. Borrowing Brown and Duguid’s
Activity theory is a general conceptual
Napier University
well-known metaphor [5], we can say that if it
approach, rather than a highly predictive
Department of
is difficult to grapple with the “whale” of con-
theory. The unit of analysis in activity theory
Computing
text by trying to get a firm grip on its specific
is the activity, consisting of a subject (an indi-
Edinburgh, EH14 1DJ,
parts, let’s try a large net instead.
vidual or group), an object or motive, arti-
United Kingdom
In this paper we present a tool that is
facts, and sociocultural rules. Leont’ev [10]
c.macaulay@dcs.napier.
directly shaped by a general theoretical
made the point that we cannot pull these
ac.uk
approach—activity theory [10, 11, 18].
pieces apart without violating the very essence
Activity theory provides a broad theoretical
of human activity, just as we cannot pull apart
framework for describing the structure, devel-
sodium and chloride if we want to understand
opment, and context of human activity. In the
salt. Understanding human activity requires a
1990s, activity theory has been applied to
commitment to a complex unit of analysis.
problems of human–computer interaction by
Two basic ideas animate activity theory: (1)
an international community of scholars and
the human mind emerges, exists, and can only
practitioners [1–4, 8, 9, 12].
be understood within the context of human
Activity theory is framed by several basic
interaction with the world; and (2) this inter-
principles (explained in the next section):
action, that is, activity, is socially and cultur-
hierarchical structure of activity, object-ori-
ally determined. These ideas are elaborated in
entedness, internalization and–externaliza-
activity theory into a set of five principles as
tion, tool mediation, and development. These
follows.
general principles help orient thought and
research, but they are somewhat abstract when
Object-Orientedness
it comes to the actual business of working on
The principle of object-orientedness states
a design or performing an evaluation. To
that every activity is directed toward some-
make activity theory more useful, we have
thing that objectively exists in the world, that
developed an artifact—the Activity
is, an object. For example, a computer pro-
Checklist—that makes concrete the concep-
gram is an object of a programmer’s activity.
28
i n t e r a c t i o n s . . . j u l y + a u g u s t 1 9 9 9

Methods&Tools p27-39 6/7/99 6:37 AM Page 29
Human activity can be oriented toward two
rent situations. Actions transform into
types of objects: things and people [10]. The
operations when they become routinized
notion of an object is not limited in activity
and unconscious with practice. When
theory to the physical, chemical, and biolog-
learning to drive a car, the shifting of the
ical properties of entities. Socially and cultur-
gears is an action with an explicit goal that
ally determined properties are also objective
must be consciously attended to. Later,
properties that can be studied with objective
shifting gears becomes operational and
methods. For example, the intended purposes
“can no longer be picked out as a special
and ways of using artifacts can be objectively
goal-directed process: its goal is not picked
studied.
out and discerned by the driver” [10].
Conversely, an operation can become an
Hierarchical Structure of Activity
action when “conditions impede an
According to Leont’ev [11], interaction
action’s execution through previously
METHODS & TOOLS
between human beings and the world is orga-
formed operations” [10]. For example, if
COLUMN EDITORS
nized into functionally subordinated hierar-
one’s mail program ceases to work, one
Michael Muller
chical levels. Leont’ev differentiated among
continues to send mail by substituting
Lotus Development Corp.
three levels: activities, actions, and operations.
another mailer, but it is now necessary to
55 Cambridge Parkway
Activities are undertaken in order to fulfill
pay conscious attention to using an unfa-
Cambridge, MA 02142 USA
motives. Motives can be considered top-level
miliar set of commands. This dynamic
+1-617-693-4235
objectives that are not subordinated to any
movement up and down the hierarchy dis-
fax: +1-617-693-1407
other objectives. Behind a motive “… there
tinguishes the activity theory hierarchy
mullerm@acm.org
always stands a need or a desire, to which [the
from static models such as GOMS.
activity] always answers” [10]. People may or
Finn Kensing
may not be consciously aware of their motives.
Internalization and Externalization
Computer Science
Actions are goal-directed processes that
Activity theory differentiates between
Roskilde University
must be carried out to fulfill a motive. For
internal and external activities. The tradi-
P.O. Box 260
instance, a programmer may write a utility
tional notion of mental processes (such as
DK-4000 Roskilde
program needed to make his larger program
in cognitive science) corresponds to
Denmark
work efficiently. The larger program itself
internal activities. Activity theory empha-
+45-4675-7781-2548
might be an action with respect to a motive
sizes that internal activities cannot be
fax: +45-4674-3072
such as getting ahead at work. Actions are con-
understood if they are analyzed separately,
kensing@dat.ruc.dk
scious; people are aware of their goals.
in isolation from external activities, because
Goals can be broken into lower level goals,
it is the constant transformation between
which, in turn, can have lower level goals,
external and internal that is the very basis
much like the concept of goals and subgoals in
of human cognition and activity.
artificial intelligence (AI) and other traditions.
Internalization is the trans-
For example, writing a utility program might
formation of external activi-
involve talking to another programmer about
ties into internal ones.
how she solved a similar problem, which
Activity theory empha-
might involve scheduling a time to talk,
sizes that not only do
opening an electronic calendar, and so forth.
mental representations get
Actions are similar to what are often referred
placed in someone’s head,
to in the human–computer interaction litera-
but the holistic activity,
ture as tasks [e.g., 15].
including motor activity and
Moving down the hierarchy of actions we
the use of artifacts, is crucial
cross the border between conscious and auto-
for internalization. For example,
matic processes. Functional subunits of
learning to calculate may involve
actions, which are carried out automatically,
counting on the fingers in the
are operations. Operations do not have their
early stages of learning simple
own goals; rather they adjust actions to cur-
arithmetic. Once the arith-
i n t e r a c t i o n s . . . j u l y + a u g u s t 1 9 9 9
29

Methods&Tools p27-39 6/7/99 6:37 AM Page 30
metic is internalized, the calculations can be
tion can make an action more efficient
performed in the head without external aids.
because external components, such as per-
Internalization provides a means for people to
forming calculations in the head, are omitted.
try potential interactions with reality without
Externalization transforms internal activi-
performing actual manipulation with real
ties into external ones. Externalization is often
objects (mental simulations, imaginings, con-
necessary when an internalized action needs to
sidering alternative plans, and so forth).
be repaired, or scaled, such as when a calcula-
Therefore, internalization can help identify
tion is not coming out right when done men-
optimal actions before actually performing an
tally or is too large to perform without pencil
action externally. In some cases, internaliza-
and paper or calculator (or some external arti-
T h e C h e c k l i s t i n t h e F i e l d
Catriona Macaulay
As a tool for thinking, the Checklist lends itself to many situations and uses. In this section I illustrate one
such situation—the domain investigation—with a personal account of my experiences using the Checklist.
For some time now, the need to “contextualize” the design of computer systems has been recognized [1].
Context is of course a notoriously slippery term, and contextualizing design can mean anything from simply
taking into account the physical environment in which a system is to be used to developing richly detailed
accounts of how people do the things we design new artifacts to support. Ethnographic techniques (see [3]
for an introduction) have become firmly established as one way of gathering contextual information. The
uses of ethnography within design settings have been described as a continuum, ranging from requirements
gathering tied to a particular development project, to opening up a broad domain such as information gath-
ering in order to contribute to our currently limited understanding about fundamental tasks. [4]
My field site was a UK national daily newspaper. I had gone there to explore what ‘information gathering’
meant in the context of journalism. And I was doing this for a very explicit purpose, that of informing the
design of future technologies to support such activities. Like many ethnographers, having made the decision
to go into the field I was unsure about what to do when I got there. To complicate matters, I came from a
background in computing and human-computer interaction studies and therefore was taking a particular
information technology-biased set of preconceptions and inclinations into the field with me.
These issues, my natural inclinations towards theory, and my inexperience as a fieldworker all led me to
look for some kind of theoretical scaffolding. Activity Theory (AT) seemed a good choice. AT, I reasoned, had
been investigated within HCI and CSCW circles for some time. [2] It seemed to offer hope for bridging the
field-design gap by providing a set of concepts relevant to both AT researchers and designers. And activity
theory provided a particularly rich set of insights into the relationship between artifacts and practice.
The adoption of theoretical frameworks is, of course, not without its dangers. Prior to conducting my
main study, I undertook a short pilot study at a small community organization. Very quickly during this
period I felt overwhelmed by my attempts to orient my field experiences around activity theory issues, and
I eventually abandoned the attempt. It was at this point that I fortuitously discovered the Checklist. Now I
had something tangible I could use. It gave me a quick way of relating experiences in the field to AT con-
cepts. It helped me think about the kinds of data I wanted to gather, and the kinds of questions I wanted to
ask. As time went on, field driven concerns came to dominate my efforts and the Checklist took more of a
back seat. When I was out of the field and reviewing my notes and transcripts, the Checklist once again
gave me an additional viewpoint on it all.
But how did I actually use it? Well, one of the Checklist’s benefits is its informality. Key concepts are illus-
trated with sample questions which suggest avenues for thought and exploration rather than formal direc-
tions. The Checklist orients without prescribing. I reduced the main section of the Checklist to A5 and kept a
copy in my fieldnotes books as an aide memoir. This proved particularly handy for the nervous neophyte
fieldworker I was. It gave me something to look at and think about in those awful moments sitting around
in the field feeling completely lost! I also had a copy stuck on my office wall which I could refer to when I
was preparing for interviews or observation sessions. During data analysis, the Checklist provided a
i n t e r a c t i o n s . . . j u l y + a u g u s t 1 9 9 9
30

Methods&Tools p27-39 6/7/99 6:37 AM Page 31
fact). Externalization is also important when
tools shape the way human beings interact
collaboration between several people requires
with reality. Shaping external activities results
their activities to be performed externally in
in shaping internal ones. Second, tools usually
order to be coordinated.
reflect the experience of other people who
tried to solve similar problems before and
Mediation
invented or modified the tool to make it more
Activity theory’s emphasis on social factors
efficient and useful. This experience is accu-
and on the interaction between people and
mulated in (1) the structural properties of
their environments explains why the principle
tools (shape, size, material) and (2) the knowl-
of tool mediation plays a central role. First,
edge of how the tool should be used. Point 2
resource for deriving additional codewords for my data analysis work.
Just having a copy visible when I was working that I could occasionally look up at was helpful. The sample
questions were particularly useful. For example, one day I caught sight of one of the sample questions
under the Environment column:
Are concepts and vocabulary of the system consistent with the concepts and vocabulary of the domain?
I suddenly realized that whilst in computing people talk about “information” all the time, in journalism
people talk about “sources.” Computing people design systems primarily to help people find information,
but “information” is often treated by designers simplistically. Journalists, on the other hand, are more inter-
ested in the sources of information than in the information itself. The challenge is finding a source for infor-
mation about something within an extremely limited timeframe. Subjective judgements about the relevance
of a piece of information, then, are made largely in relation to judgements about the source. This led me to
the realization that sources can be seen as a very particular kind of artifact within journalistic information
gathering, and that they have largely been overlooked by designers of information gathering systems.
During the early stages of my study, the sample questions helped me understand the specific issues the
Checklist deals with. Later, as my understanding grew, I turned more to the issues in the Checklist rather
than the sample questions. Later still, I found myself developing my own sample questions, questions I now
carry with me into my next study.
Activity theory and the Checklist also proved a useful counter to my natural inclination to cling to the
familiar—to obvious technological artifacts. Entering the world of ethnographic fieldwork from a computing
background, one can easily become over-focussed on high-tech devices, or on “information” in a simplistic
sense. During my first forays into the field, I was so focussed on what I thought the obvious constituents of
information gathering activity would be, that I completely failed to recognize the importance of sources. It
was this kind of benefit from the Checklist that I most valued. The Checklist was a tool for reflexivity,
helping me in my attempts to maintain an awareness of where my own instinctive concerns and interests
were closing me off from those of the people I was studying.
In summary then, the Checklist became a valuable aide memoir and a tool for reflexivity. Although the
Checklist as presented here does not explicitly draw attention to its reflexive uses, this was clearly some-
thing of particular benefit to more broadly scoped fieldwork such as mine. For the theoretically-oriented
fieldworker, the Checklist provides a flexible and non-prescriptive way of maintaining an awareness of
potentially relevant aspects of AT to design concerns. Of course it does not do away with the need to
engage with the ideas behind activity theory more broadly, but it certainly helps kick-start the process.
References
1. Clarke, S. (1997). Encouraging the Effective Use of Contextual Information in Design. Unpublished PhD, University of Glasgow,
Glasgow, Scotland.
2. Draper, S. (1992). Activity Theory: The New Direction for HCI? International Journal for Man-Machine Studies, 37.
3. Hammersley, M., & Atkinson, P. (1995). Ethnography: Principles in Practice. (2nd ed.). London: Routledge.
4. Whittaker, S., Terveen, L. and Nardi, B. Let’s stop pushing the envelope and start addressing it. Submitted to TOCHI.
i n t e r a c t i o n s . . . j u l y + a u g u s t 1 9 9 9
31

Methods&Tools p27-39 6/7/99 6:37 AM Page 32
is critical for activity theory. Many theories
aspects of the whole activity. A systematic
discuss Point 1 (such as the idea of affor-
application of any of these principles makes it
dances, Latourian notions of tool prescrip-
eventually necessary to involve all the others.
tions, and so forth). Activity theory
For instance, understanding the hierarchical
emphasizes that a tool comes fully into being
structure of an activity requires an analysis of
when it is used and that knowing how to use it
its object or motive, as well as developmental
is a crucial part of the tool. So, the use of tools
transformations between actions and opera-
is an evolutionary accumulation and transmis-
tions and between internal and external com-
sion of social knowledge, which influences the
ponents. The latter, in turn, can critically
nature of not only external behavior but also
depend on the tools used in the activity.
the mental functioning of individuals.
The concept of tool in activity theory is
Activity Checklist
broad and embraces both technical tools,
As mentioned earlier, activity theory does not
which are intended to manipulate physical
provide ready-made solutions that can be
objects (e.g., a hammer), and psychological
directly applied to specific problems. We see
tools, which are used by human beings to
its main potential in supporting researchers
influence other people or themselves (e.g., the
and designers in their own search for solu-
multiplication table or a calendar).
tions, in particular, by helping them to ask
meaningful questions. To make such an appli-
Development
cation of activity theory more practical, we
Finally, activity theory requires that human
introduce an analytical tool, the Activity
interaction with reality be analyzed in the con-
Checklist.
text of development. Activity theory sees all
The Activity Checklist is intended to be
practice as being reformed and shaped by his-
used at early phases of system design or for
torical development. It is important to under-
evaluating existing systems. Accordingly, there
stand how tools are used not in a
are two slightly different versions of the
single instant of trying
Checklist, the “evaluation version” and the
them out in a labora-
“design version.” Both versions are used as
tory (for example) but
organized sets of items covering the contex-
as usage unfolds over
tual factors that can potentially influence the
time. In that time, devel-
use of a computer technology in real-life set-
opment may occur
tings. It is assumed that the Checklist can
making the tool more
help to identify the most important issues, for
useful and efficient than
instance, potential trouble spots, that
might be seen in a single
designers can address.
observation. In activity
Having two versions of the Checklist
theory, development is
implies a commitment to the study of actual
thus not only an object of
use as a critical part of design. Researchers
study, it is also a general
such as Bannon [1] have made the useful
research methodology. That is
point that design and use are two sides of the
why a basic research method in activity
same coin. Still, a design must begin some-
theory is the formative experiment which
where, and it is helpful to have guidance in
combines active participation with moni-
the earliest stages of brainstorming and cre-
toring of the developmental changes in the
ative imagining of how a technology might
object of study.
come into being.
The Checklist covers a large space. It is
Integration of the Principles
intended to be used first by examining the
These basic principles of activity theory
whole space for areas of interest, then focusing
should be considered an integrated system,
on the identified areas of interest in as much
because they are associated with various
depth as possible. The general strategy, then,
32
i n t e r a c t i o n s . . . j u l y + a u g u s t 1 9 9 9

Methods&Tools p27-39 6/7/99 6:37 AM Page 33
is breadth-first consideration of the relevant
the Checklist seems to be more effective appli-
areas of context enumerated in the Checklist,
cation of a number of already established
followed by a “drilling down” into specific
methods and techniques. For instance, the
areas that should yield rich results given the
Checklist can help identify the most relevant
tools and problems at hand.
issues to be covered in an interview or to
The structure of the Checklist reflects the
make sure important problems are not
five basic principles of activity theory. Since
overlooked in a discussion of empir-
the Checklist is intended to be applied in ana-
ical data collected in an observa-
lyzing how people use (or will use) a computer
tional study.
technology, the principle of tool mediation is
Second, the linear struc-
strongly emphasized. This principle has been
ture of the Checklist does
applied throughout the Checklist and system-
not imply that it should be
atically combined with the other four princi-
used linearly, by focusing
ples. It results in four sections corresponding
on isolated items one by
to four main perspectives on the use of the
one and ignoring the rest
“target technology” to be evaluated or
of the Checklist.
designed:
Instead, practitioners
1. Means and ends—the extent to which
using the tool should
the technology facilitates and constrains
look for patterns of related
the attainment of users’ goals and the
items, even if these items
impact of the technology on provoking or
belong to different sections.
resolving conflicts between different goals.
Third, in order to use the tool
2. Social and physical aspects of the envi-
effectively, practitioners should familiarize
ronment—integration of target tech-
themselves with the Checklist and even try to
nology with requirements, tools,
internalize it. We recommend that practi-
resources, and social rules of the environ-
tioners follow the items in the Checklist
ment.
repeatedly at various phases of design or eval-
3. Learning, cognition, and articulation
uation. A quick initial run should identify the
internal versus external components of
most important potential trouble spots and
activity and support of their mutual trans-
filter out the rest. Further runs may result in
formations with target technology.
finding patterns, revising previously made
4. Development—developmental transfor-
judgments about the importance or unimpor-
mation of the foregoing components as a
tance of certain issues, and formulating
whole.
requests for more information, if necessary.
Taken together, these sections cover various
Fourth, it should be noted that every tool
aspects of how the target technology supports,
is used for some purpose, and the Checklist is
or is intended to support, human actions
no exception. Therefore, potential users of the
(“target actions”). See the Checklist in the
Checklist should clearly understand why they
Appendix to this paper.
are using the tool. Such understanding can
help focus on relevant items and ignore irrele-
Using the Checklist
vant ones. Also, such understanding is neces-
According to our experience of using the
sary for successful incorporation of
Checklist and teaching other people how to
conclusions, judgments, and ideas related to
use it, there are several points to remember
individual items into more general notions
when trying to apply the tool in a specific pro-
relevant to design or evaluation of the system
ject.
as a whole.
First, the Checklist is supposed to be used
not as the only basis for system design or eval-
Apple Data Detectors: An Example of
uation, but in combination with other tech-
Using the Checklist
niques. One of the main advantages of using
The design of Apple Data Detectors, a multi-
i n t e r a c t i o n s . . . j u l y + a u g u s t 1 9 9 9
33

Methods&Tools p27-39 6/7/99 6:37 AM Page 34
purpose intelligent agent technology for ana-
ferent because the emphasis was not the users’
lyzing and taking action on structured data
wider context as it was in Apple Data
[14], is an example of using the Checklist to
Detectors. Apple Data Detectors allows for
go beyond the narrow scope of many design
end-user development of structures, scripting
projects. Apple Data Detectors recognizes
of actions, mixing and matching recognizers
structured data such as URLs, e-mail
and actions, and composite structures (see
addresses, postal addresses, ISBN numbers,
Table 1). It provides flexibility and a growth
and stock symbols. Using “structure detec-
path for users.
tion” technology [17], a detector is written to
Most designers will have to be concerned
analyze structured data. The detector is then
with the Means/ends column of the Checklist.
paired with an action such as “Open URL” or
In Apple Data Detectors we studied potential
“Create e-mail message.” Apple Data
uses of Data Detectors and found that for the
Detectors works in any text; applications do
technology to be useful, users need composite
not need to be modified to use it.
structures such as postal addresses. It is con-
In the design of Apple Data Detectors, we
siderably more difficult to write a parser that
were concerned with the Learning/develop-
handles composite structures (e.g., an address
ment areas of the Checklist. We devoted many
is composed of a name, street, city, etc., each
resources to considering how end users would
of which is an atomic structure). But our users
go from simple use of the tool involving only
can select an address (with the mouse), and
accessing structures supplied by Apple or
Apple Data Detectors will recognize it and
third-party developers to programming their
take a prespecified action such as adding the
own new agents. As we considered the prin-
address to the user’s address book, putting
ciple of development from the beginning of our
each field of the address in the appropriate
research, we were able to create an architecture
place in the address book.
that supports end-user programming [12]. We
We also gave careful thought at the outset
paid special attention to the first four areas in
of the project to our criteria for success and
the Development column. To reiterate the
failure (the Design section of the Means/ends
point made earlier, the Checklist can be used
column). Our criteria were that the tech-
to scope out a large possible space of potential
nology be useful for Apple customers and that
areas of interest and then narrow down to spe-
developers be able to use it painlessly. (Third-
cific areas to actively pursue. The Checklist is
party developers are developing the structures
useful in reminding developers of a larger
and actions that work with their applications.)
space that gets beyond the details of user
We thus decided not to use OpenDoc, even
interface mechanisms and leads systematically
though there was some pressure to do so. But
into many areas of the context of use that may
the developers’ experience would have been
provide inspiration for interesting designs.
much more difficult. As it turned out, this was
A research project at another lab used
the right decision more than we knew at the
structure detection technology much as we
time, because OpenDoc was eventually put on
did [16]. But the prototype looked quite dif-
the corporate back burner. It was the explicit
Table 1 A COMPARISON OF TWO USES OF STRUCTURE DETECTION TECHNOLOGY
Intel Selection Recognition Agent
Apple Data Detectors
No scripting
(C behind an API)
AppleScript
No path to end-user modification
Script editor
End-user modification
Recognizer/action pairs bound together
Separate recognizers, actions
No composite structures
Composite structures
34
i n t e r a c t i o n s . . . j u l y + a u g u s t 1 9 9 9

Methods&Tools p27-39 6/7/99 6:37 AM Page 35
attention to a firm set of design criteria that
multiyear study, but that is not how it is
helped us weather that storm.
intended to be employed. For most uses of the
Checklist, users should first do a “quick-and-
Conclusion
dirty” perusal of the areas represented in the
As mentioned earlier, the Activity Checklist is
Checklist that are likely to be trouble-
not the only attempt to deal with context in
some or interesting (or
the field of HCI, and it is not intended as a
both) in a specific design
substitute for other approaches. From our
or evaluation. Then,
point of view, the Checklist can be most suc-
once those areas have
cessfully used together with other tools and
been identified, they
techniques to efficiently address issues of con-
can be explored
text.
more deeply.
For instance, task analysis [6] places a heavy
The breadth of
emphasis on the Means/ends dimension of
coverage in the
context, whereas environment and, especially,
Checklist will help to
learning and development are underrepre-
ensure that designers do
sented. Using the terminology of activity
not miss areas that might be
theory, we could say that task analysis gives a
important for understanding the
thorough description of individual actions,
tool they are working on.
whereas the higher levels of activity and inter-
relations between actions receive less attention.
Acknowledgments
Contextual design [7], conversely, provides
Many thanks to Helen Hasan,
an elaborated set of concepts and techniques
Mark Spasser, Clay Spinuzzi, and John
for describing the environment (as in the
Waterworth for their helpful comments on an
“Environment” section of the Checklist). It
earlier version of the paper.
also supports identifying users’ tasks
(although, in our opinion, not to the extent to
References
which it addresses the structure and func-
1. Bannon, L. From human factors to human actors:
tioning of the environment) but is less focused
The role of psychology and human–computer interac-
on learning and development. Finally,
tion studies in system design. In Design at Work:
Development is a major concern within par-
Cooperative Design of Computer Systems, J. Greenbaum
ticipatory design approaches [e.g., 3], but
and M. Kyng, eds., Lawrence Erlbaum, Hillsdale, NJ,
identifying task structures does not have a
1991.
high priority there.
2. Bødker, S. Through the Interface: A Human Activity
Therefore, each of the existing empirically
Approach to User Interface Design. Lawrence Erlbaum,
driven approaches to context has its strong
Hillsdale, NJ, 1991.
points. From our point of view, the main
3. Bødker, S., Knudsen, J., Kyng, M., Ehn, P., and
advantage of the Activity Checklist is that it is
Madsen, K.H. Computer Support For Cooperative
a general framework that can be used to (1)
Design. In Proceedings of ACM CSCW’88 Conference on
provide a preliminary overview of potentially
Computer-Supported Cooperative Work (Portland, OR,
relevant contextual factors, (2) select appro-
1988).
priate tools for further exploration, and (3)
4. Bourke, I., Verenikina I., and Gould, E. Interacting
evaluate limitations of those tools. In other
With Proprietary Software Users: An Application for
words, the Checklist can help to leverage the
Activity Theory? In Proceedings East-West International
various strengths of empirically based
Conference on Human-Computer Interaction (Moscow,
approaches.
August 3–7, 1993,). ICSTI, Moscow.
The fact that the Checklist is comprehen-
5. Brown, J., and Duguid, P. Borderline issues: Social
sive and wide-ranging should not mislead its
and material aspects of design. Human-Computer
potential users. It would be impossible to
Interaction 9, 1 (1994).
investigate all the areas it covers without a
6. Dix, A., Finlay, J. Abowd, G., and Beale, R. Human-
i n t e r a c t i o n s . . . j u l y + a u g u s t 1 9 9 9
35

Methods&Tools p27-39 6/7/99 6:37 AM Page 36
Computer Interaction. Prentice Hall, London, 1993.
13. Nardi, B., ed. Context and Consciousness: Activity
7. Holtzblatt, K., and Beyer, H. Making customer-cen-
Theory and Human–Computer Interaction. MIT Press,
tered design work for teams. Communications of the
Cambridge, MA, 1996.
ACM 36, 10 (1993).
14. Nardi, B., Miller, J., and Wright, D. Collaborative,
8. Kaptelinin, V. Human–Computer Interaction in
Programmable Intelligent Agents. In Communications of
Context: The Activity Theory Perspective. In
the ACM March 1998.
PERMISSION TO MAKE DIGITAL OR
Proceedings of the East-West Human–Computer
15. Norman, D. Cognitive Artifacts. In Designing
HARD COPIES OF ALL OR PART OF THIS
WORK FOR PERSONAL OR CLASSROOM
Interaction Conference (St. Petersburg, Russia, August
Interaction: Psychology at the Human-Computer Interface,
USE IS GRANTED WITHOUT FEE
4–8). ICSTI, Moscow, 1992.
J. Carroll. ed., Cambridge University Press, Cambridge,
PROVIDED THAT COPIES ARE NOT
9. Kuutti, K. Activity theory and its applications in
1991.
MADE OR DISTRIBUTED FOR PROFIT OR
information systems research and design. In Information
16. Pandit, M., and Kalbag, S. The Selection
COMMERCIAL ADVANTAGE AND THAT
COPIES BEAR THIS NOTICE AND THE
Systems Research Arena of the 90’s, H.-E. Nissen, H.K.
Recognition Agent: Instant Access to Relevant
FULL CITATION ON THE FIRST PAGE.
Klein, and R. Hirschheim, eds., Elsevier, Amsterdam,
Information and Operations. In Proceedings of
TO COPY OTHERWISE, TO REPUBLISH,
1991.
Intelligent User Interfaces ‘97. (1997 Orland, Fl.),
TO POST ON SERVERS OR TO REDIS-
TRIBUTE TO LISTS, REQUIRES PRIOR
10. Leont’ev, A.N. Activity, Consciousness, Personality.
New York, ACM Press.
SPECIFIC PERMISSION AND/OR A FEE.
Prentice Hall, Englewood Cliffs, NJ, 1978.
17. Rus, D. and Subramanian, D. Multi-media RISSC
© ACM 1072-5220/99/0700 $5.00
11. Leont’ev, A. N. Problems of the Development of
Informatics. In Proceedings of the 2nd International
Mind. Progress, Moscow, 1981.
Conference on Information and Knowledge Management.
12. Lieberman, H., Nardi, B. and Wright, D. Training
(Washington, DC, 1993.). ACM Press, New York, pp.
Agents to Recognize Text by Example. In Proceedings of
283–294.
the International Conference on Autonomous Agents,
18. Wertsch, J. Ed., The Concept of Activity in Soviet
Seattle, April 1999.
Psychology. M. E. Sharpe, Armonk, NY, 1981.
A p p e n d i x . A c t i v i t y C h e c k l i s t
P R E A M B L E
Means/ends
Environment
Learning/cognition/
Development
(hierarchical structure
(object-orientedness)
articulation
(development)
of activity)
(externalization/
internalization)

Human beings have hier-
Human beings live in
Activities include both
Activities undergo per-
archies of goals that
the social, cultural
internal (mental) and
manent developmental
emerge from attempts to
world. They achieve
external components
transformations.
meet their needs under
their motives and goals
which can transform
Analysis of the history
current circumstances.
by active transformation
into each other.
of target activities can
Understanding the use of
of objects in their envi-
Computer systems
help to reveal the main
any technology should
ronments. This section
should support both
factors influencing the
start with identifying the
of the checklist identi-
internalization of new
development. Analysis
goals of target actions,
fies the objects involved
ways of action and artic-
of potential changes in
which are relatively
in target activities and
ulation of mental
the environment can
explicit, and then
constitutes the environ-
processes, when neces-
help to anticipate their
extending the scope of
ment of the use of
sary, to facilitate
effect on the structure
analysis both “up” (to
target technology.
problem solving and
of target activities.
higher-level actions and
social coordination.
activities) and “down”
(to lower level actions
and operations).
36
i n t e r a c t i o n s . . . j u l y + a u g u s t 1 9 9 9

Methods&Tools p27-39 6/7/99 6:37 AM Page 37
E V A L U A T I O N V E R S I O N
Means/ends
Environment
Learning/cognition/
Development
articulation
People who use the
Role of target technology
Components of target
Use of target technology
target technology
in producing the out-
actions that are to be
at various stages of
Goals and subgoals of the
comes of target actions
internalized
target action “life
target actions (target
Tools, other than target
Knowledge about target
cycles”—from goal set-
goals)
technology, available to
technology that resides
ting to outcomes
Criteria for success or
users
in the environment and
Effect of implementation
failure of achieving
Integration of target
the way this knowledge
of target technology on
target goals
technology with other
is distributed and
the structure of target
Decomposition of target
tools
accessed
actions
goals into subgoals
Access to tools and mate-
Time and effort necessary
New higher-level goals
Setting of target goals
rials necessary to per-
to master new opera-
that became attainable
and subgoals
form target actions
tions
after the technology
Potential conflicts
Tools and materials
Self-monitoring and
had been implemented
between target goals
shared between several
reflection through
Users’ attitudes toward
Potential conflicts
users
externalization
target technology (e.g.,
between target goals
Spatial layout and tem-
Use of target technology
resistance) and changes
and goals associated
poral organization of
for simulating target
over time
with other technologies
the working environ-
actions before their
Dynamics of potential
and activities
ment.
actual implementation
conflicts between
Resolution of conflicts
Division of labor,
Support of problem articu-
target actions and
between various goals
including synchronous
lation and help request
higher-level goals
Integration of individual
and asynchronous dis-
in case of breakdowns
Anticipated changes in
target actions and
tribution of work
Strategies and procedures
the environment and
other actions into
between different loca-
of providing help to
the level of activity they
higher-level actions
tions
other users of target
directly influence (oper-
Constraints imposed by
Rules, norms, and proce-
technology
ations, actions, or activi-
higher-level goals on
dures regulating social
Coordination of individual
ties)
the choice and use of
interactions and coordi-
and group activities
target technology
nation related to the
through externalization
Alternative ways to attain
use of target tech-
Use of shared representa-
target goals through
nology
tion to support collabo-
lower-level goals.
rative work
Troubleshooting strate-
Individual contributions
gies and techniques
to shared resources of
Support of mutual trans-
group or organization
formations between
actions and operations
i n t e r a c t i o n s . . . j u l y + a u g u s t 1 9 9 9
37

Methods&Tools p27-39 6/7/99 6:37 AM Page 38
D E S I G N V E R S I O N
Means/ends
Environment
Learning/cognition/
Development
articulation
People who use the
U
Role of existing technology
Components of target
Use of tools at various
target technology
in producing the out-
actions that are to be
stages of target action
S
Goals and subgoals of the
comes of target actions
internalized
“life cycles”—from goal
E
target actions (target
Tools, available to users
Time and effort necessary
setting to outcomes
goals)
Integration of target
to learn how to use
Transformation of existing
Criteria for success or
technology with other
existing technology
activities into future
failure of achieving
tools
Self-monitoring and
activities supported
target goals
Access to tools and mate-
reflection through
with the system
Decomposition of target
rials necessary to per-
externalization
History of implementa-
goals into subgoals
form target actions
Possibilities for simulating
tion of new technolo-
Setting of target goals
Tools and materials
target actions before
gies to support target
and subgoals
shared between several
their actual implemen-
actions
Potential conflicts
users
tation.
Anticipated changes in
between target goals
Spatial layout and tem-
Support of problem articu-
the environment and
Potential conflicts
poral organization of
lation and help request
the level of activity they
between target goals
the working environ-
in case of breakdowns
directly influence (oper-
and goals associated
ment.
Strategies and procedures
ations, actions, or activi-
with other technologies
Division of labor,
of providing help to
ties)
and activities
including synchronous
collegues and collabo-
Anticipated changes of
Resolution of conflicts
and asynchronous dis-
rators
target actions after new
between various goals
tribution of work
Coordination of individual
technology is imple-
Integration of individual
between different loca-
and group activities
mented
target actions and
tions
through externalization
other actions into
Rules, norms, and proce-
Use of shared representa-
higher-level actions
dures regulating social
tion to support collabo-
Constraints imposed by
interactions and coordi-
rative work
higher-level goals on
nation related to target
the choice and use of
actions
target technology
Alternative ways to attain
target goals through
lower-level goals.
Troubleshooting strate-
gies and techniques
Support of mutual trans-
formations between
actions and operations
Goal that can be changed
or modified, and goals
that have to remain
after new technology is
implemented
38
i n t e r a c t i o n s . . . j u l y + a u g u s t 1 9 9 9

Methods&Tools p27-39 6/7/99 6:37 AM Page 39
D E S I G N V E R S I O N
Means/ends
Environment
Learning/cognition/
Development
articulation
Parties involved in the
Resources available to the
Representations of design
Anticipated changes in
D
process of design
parties involved in
that support coordina-
the requirements to the
E
Goals of designing a new
design of the system
tion between the par-
system
system
Rules, norms, and proce-
ties
S
Criteria of success or
dures regulating inter-
Mutual learning of the
I
failure of design
action between the
content of the work
Potential conflicts
parties
(designers) and possibil-
G
between goals of
ities and limitations of
N
design and other goals
technology (users)
(e.g., stability of the
organization, mini-
mizing expenses)
S A M P L E Q U E S T I O N S
Means/ends
Environment
Learning/cognition/
Development
articulation
Are all target actions
Are concepts and vocabu-
Is the whole “action life-
What are the conse-
actually supported?
lary of the system consis-
cycle,” from goal setting
quences of implementing
tent with the concepts
to the final outcome,
the target technology on
Is there any functionality
and vocabulary of the
taken into account and/or
target actions? Did
of the system that is not
domain?
supported?
expected benefits actually
actually used? If yes,
take place?
which actions were
Is target technology con-
Does the system help to
intended to be supported
sidered an important part
avoid unnecessary
Did users have enough
with this functionality?
of work activities?
learning?
experience with the
How do users perform
system at the time of
these actions?
Are computer resources
Is externally distributed
evaluation?
necessary to produce a
knowledge easily acces-
Are there actions, other
certain outcome inte-
sible when necessary?
Did the system require a
than target actions, that
grated with each other?
large investment of time
are not supported, but
Does the system provide
and effort in learning
users obviously need such
Is target technology inte-
representations of user’s
how to use it?
support?
grated with other tools
activities that can help in
and materials?
goal setting and self-eval-
Did the system show
Are there conflicts
uation?
increasing or decreasing
between different goals
Are characteristics of
benefits over the process
of the user? If yes, what
target technology consis-
Does the system provide
of its use?
are the current trade-offs
tent with the nature of
problem representations
and rules or procedures
the environment (e.g.,
in case of breakdowns
Are users’ attitudes
for resolving the conflicts?
central office work vs.
that can be used to find a
toward the system
teleworking)?
solution or formulate a
becoming more or less
What are the basic limita-
request for help?
positive?
tions of the current tech-
nology?
Are there external repre-
Are there negative or
sentations of the user’s
positive side-effects asso-
Is it necessary for the user
activities that can be used
ciated with the use of the
to constantly switch
by others as clues for
system?
between different actions
coordinating their activi-
or activities? If yes, are
ties within the framework
there “emergency exits”
of group or organization?
which support painless
transition between actions
and activities, and, if neces-
sary, returning to previous
states, actions, or activities?
i n t e r a c t i o n s . . . j u l y + a u g u s t 1 9 9 9
39