A Design Theory for Cognitive Work°ow Systems
Jarkko Hyysalo
*
,
†
, Markku Oivo
*
,
‡
and Pasi Kuvaja
*
,
§
*
Faculty of Information Technology and Electrical Engineering, M3S
University of Oulu, P. O. Box 8000
FI-90014 Oulu, Finland
†
jarkko.hyysalo@oulu.fi
‡
markku.oivo@oulu.fi
§
pasi.kuvaja@oulu.fi
Received 24 October 2014
Revised 22 November 2014
Accepted 10 February 2016
Published
This paper addresses the design problem of providing cognitive support for work°ow systems in
software development. Software development is demanding knowledge work that requires cre-
ativity and adaptability to changing requirements and situations. This type of work involves
cognitive actions that require substantial support in several forms in order to address needs such
as collaboration, communication, knowledge management, awareness and transparency, and
the coordination and structuring of the development processes. The literature and our empirical
results show that there is a lack of cognitive support in current work°ow models. Hence, we
identify the need for a design theory for cognitive work°ow systems (CWS). In this paper, such a
theory is presented. The proposed design theory for CWS is validated through an action re-
search intervention. This design theory has important implications from both research and
practical perspectives. The results will help developers in their daily work, enhance the e±ciency
of the development processes, and facilitate decision-making activities.
Keywords: Cognition support; cognitive work; design theory; knowledge work; software devel-
opment; work°ow; work support.
1. Introduction
Software development is increasingly challenging and intellectually demanding cre-
ative knowledge work [1]. It is a human- and knowledge-intensive activity in which
managing knowledge is paramount [2]. Such information-intensive knowledge work is
primarily a cognitive activity [3] based on the worker's internal mental processes
rather than on physical labor. Cogniti ve perspectives are fundamental factors for
successful collaboration [4]. Thus, knowledge work and cognitive aspects should be
†
Corresponding author.
International Journal of Software Engineering
and Knowledge Engineering
Vol. 27, No. 1 (2017) 1750006 (27 pages)
#
.
c
World Scienti¯c Publishing Company
DOI: 10.1142/S0218194017500061
November 8, 2016 3:53:39pm WSPC/117-IJSEKE 1750006
2
nd
R
ea
di
ng
1
supported properly in order to obtain good results. Software systems are usually
developed as a multidisciplinary e®ort [5], typicall y in collaboration with several
types of stakeholders such as engineers, industrial designers, and marketing per-
sonnel [6, 7].
Work°ows are used to support these collaborative processes. Here, we d e¯ne a
work°ow as a sequence of working steps or logicall y- re l a t ed t asks, including the use
of resources to achieve a common goal !!! transforming a set of inputs into outputs
that provide value for stakeholders. However, it is not easy to plan the work in
detail beforehand [8], as software processes are often complicated in nature; they
involve a large number of tasks, performers, and coordinat ion constraints. More-
over, cognitive work and cooperation are not properly supported in current
work°ow models [9, 10].
A practical solution is needed for supporting cognitive work in collaborative de-
velopment. Hence, we need to design a work°ow that addresses the above-mentioned
issues. As this is a design issue, we claim that the design theory approach proposed in
[11] is an e®ective approach to address it. Bringing forth a design theory for a
cognitive work°ow system (CWS) is the main contribution of this work.
Building a work°ow that supports cognitive work !!! intellectually demanding
and creative knowledge work !!! requires both the development of a formal meth-
odology that can help model the knowledge °ows and an extension of current
work°ow technologies to handle information that is characterized by dynamic
changes, requirements for innovative problem solving, and cognitive processing. The
framework and formalisms, we propose in this paper serve as the foundation for such
a methodology. Design theory has proven to be successful for the purposes of de-
scribing improved design processes [12]. We propose that the kernel and/or justi¯-
catory theories and extensive case studies resulting in the design theory provide a
solid basis for prescribing the principles and requirements for designing and con-
structing an e®ective artifact.
The remainder of this paper is organized as follows. Section 2 discusses related
studies, work°ow modeling, and design theory and states the motivation for this
research. Section 3 presents the research process and provides the theoretical
framework for this research in the form of kernel theories. Further, it describes the
results of empirical work. Section 4 evaluates how the implementation of the six
design principles addresses the identi ¯ed issues by analyzing the research ¯ndings,
also including a comparison to existing approaches. Section 5 discusses the results,
implications, and future research directions and concludes the paper.
2. Theoretical Background
This section presents the theoretical concepts related to work°ow modeling in the
context of knowledge work, discusses related studies, and states the motivation for
this research. In addition, the design theory approach is brie°y described.
2 J. Hyysalo, M. Oivo & P. Kuvaja
November 8, 2016 3:53:40pm WSPC/117-IJSEKE 1750006
2
nd
R
ea
di
ng
2.1. Cognitive work support in work°ows
Software development involves several challenges related to individual and team
cognition, such as complex decision-making, innovative problem solving, handling of
vast amounts of information, building of a shared understanding, and information
and knowledge sharing [13]. The greater the °ow of knowledge, knowledge sharing,
and learning between di®erent organizational parts and external agents, the greater
are the opportunities for knowledge generation [14
–
16]. However, one of the pro-
blems in knowledge work is the number of disconnected tools, that is, tools that are
disconnected not only from other tools but also from the work processes; moreover,
knowledge is often disconnected from the context [17]. Thus, a wor k°ow that sup-
ports knowledge sharing and enables the use of cognitive skills is required. The
work°ow must also be designed such that it addresses di®erent views, including those
of business analysts and software developers [18], and connects the knowledge to
work, processes, and the context of work.
In this study, cognitive activities and processes include, for example, the following
aspects of knowledge work: making observ ations; reasoning; processing information;
learning, understanding, and remembering information content, using information
systems and conceptualizing.
Support for cognitive work is crucial when the work concerns abstract matters
and knowledge, and the following challenges are identi¯ed: knowledge is not easily
transferred unless it is made explicit, knowledge elements are context-speci¯c, and
cooperation is needed because of humans' cognitive limitations !!! one does not know
everything [19]. Cognitive skills are also required to respond to changes, that is,
creativity and human problem-solving skills are required to deal with changes and
unexpected events [20].
It has been proposed that the purpose of collaborative software development
environments is to facilitate and nurture developers' creative knowledge processes
[21]. For this purpose, for example, companies implement wor k°ows to help manage
processes, transfer work and data from one worker to anoth er, and establish a logical
order for task implementation. However, it has been acknowledged that traditional
work°ow approaches are too static and do not address the changes and unexpected
events that inevitably occur in creative knowledge work; moreover, work°ow models
lack cognitive support [9, 22
–
25]. Nor are cognitive work and cooperation properly
supported in current work°ow models [9, 10].
Hence , there is a need for work°ow design principles to provide guidance on how
to model and implement a CWS that supports software development processes in
which uncertainty prevails and how know ledge °ows can be utilized to create a
solid basis fo r deci sion-making activities. We suggest that most cognitive chal-
lenges can be tackled with properly designed cognitive work°ows that address
collaboration and communication practices, implicit and explicit knowledge man-
agement, awareness and transparency, and processes supporting coordination
and negotiation.
A Design Theory for Cognitive Work°ow Systems 3
November 8, 2016 3:53:40pm WSPC/117-IJSEKE 1750006
2
nd
R
ea
di
ng
Abstracting and synthesizing from related works, we identi¯ed a set of char-
acteristics of software development, which we classi¯ed into six categories: cognitive
work support, collaborative work, communication, knowledge management, aware-
ness and transparency, and coordination. Table 1 presen ts a brief summary of this
categorization. Then, we mapped each characteristic to a design principle for CWS
and identi¯ed its requirements relevant to a work°ow system.
Other characteristics and requirem ents could also be identi¯ed; however, this
work addressed the research problem from these six viewpoints, all of which com-
plement each other and c ontribute to de¯ning cognitive support for collaborative
work. Many other concepts and viewpoints can also be seen as relating to CWS, such
as distributed work, global software development, decision-support systems, virtual
collaboration tools, learning, and computer-supported cooperative work. H owever,
this work did not address these areas individually but, rather, focused on viewing
work°ow systems from a cognitive point of view.
2.2. Design theory approach
Design theory has proven to be successful for the purposes of describing improved
design processes [12], and there are examples of successful attempts to use the design
theory approach; see [17, 26
–
29]. This section brie°y describes the concept. For a
more detailed description, see, for example, [11, 12, 30
–
32], where the use of design
theory is discussed extensively.
Table 1. Summary of characteristics and their system support requirements.
Software development
characteristics Requirements for work°ow systems Design principles for CWS
Complex tasks, knowledge-in-
tensive processes, abstract
knowledge
Supporting understanding, shared
understanding, and reducing
cognitive load
Cognitive work support
Di®erent kinds of tasks, di®er-
ent expertize areas, distrib-
uted work, team work
Supporting collaboration, distrib-
uted interactions, and realiza-
tion of business processes
Collaborative work
Information exchange, under-
standing of each other
Creating and maintaining shared
understanding
Communication
Knowledge-intensive processes,
team knowledge, knowledge
management
Acquiring, sharing, and using
knowledge
Knowledge management
Understanding of others' activ-
ities, context, information
visualization, situation
awareness, workspace
awareness
Enabling knowledge, supporting
awareness and transparency
Awareness and transparency
Simultaneous, sequential,
distributed
Creating structure, synchroniza-
tion, integrating contributions,
decision points
Coordination
4 J. Hyysalo, M. Oivo & P. Kuvaja
November 8, 2016 3:53:40pm WSPC/117-IJSEKE 1750006
2
nd
R
ea
di
ng
Design theory describes components and their relationships for subsystems that
are then used to construct the system [11]. In other words, the \pieces" can be
designed relatively independent of one another; however, they eventually form a
complete system in which the subsystems are not completely independent. Thus, all
the pieces and the entire system must be satisfactorily constructed !!! this is the goal
of design theory. Another important aspect of design theory is that it deals with both
the product and the process of design; these two cannot be completely isolated,
because the process eventually yields the product [11].
Design theory consists of product and process aspects, and for both aspects, there
are kernel theories that describe the fundamental behavior underlying the concepts,
for example, the laws of nature. Meta-requirements are derived from the kernel
theories, and they describe a class of goals to which the theory is applicable [11].
Based on the meta-requirements, meta-design principles are drawn up to describe the
class of artifacts that are hypothesized to meet the meta-requirements. Finally, there
are testable design hypotheses for verifying whether the meta-design principles meet
the meta-requirements. However, it should be noted that context-speci¯c modi¯ca-
tions are required to accommodate domain-speci¯c aspects.
Walls et al. [11] formally speci¯ed and coined the term, \design theory," and
others have also used and modi¯ed the approach . Thus, there are several useable
ways of presenting design theories. The original approach by Walls et al. [11] has
been reviewed in literature, and there are considerations that it might initially have
been too cumbersome to use (cf. [29, 33]).
It has been argued that testable design hypotheses are unnecessary and not
essential to the theory, as design theory explains generalized solution components
with the related generalized requirements or design principles [31]. Gregor and
Jones [30] also agree, and they suggest that even without testable hypotheses and
empirical indicators, desi gn theory is su±cient to describe the idea of an artifact
that could be constru cted. Th ere are also other examples wh ere the testable hy-
potheses ar e n ot seen as necessary components of design theory; see, for example,
[29, 32].
Building the design theory in this article follows the approach presented in
Hanseth and Lyytinen [34]. Hanseth and Lyytinen also base their approach on Walls
et al. [11], but they have made small changes; for example, they have omitted the use
of testable hypotheses.
Table 2. Components of a design theory [32].
Design goals Describe the goals to which the theory applies.
A set of system features A set of artifacts hypothesized to meet the requirements.
Kernel theories Theories or justi¯catory knowledge from natural and social sciences governing
the design requirements or the processes for arriving at them.
Design principles A codi¯cation of procedures, which, when applied, increase the likelihood of
achieving a set of system features. These procedures are derived from
kernel theories.
A Design Theory for Cognitive Work°ow Systems 5
November 8, 2016 3:53:40pm WSPC/117-IJSEKE 1750006
2
nd
R
ea
di
ng
![Fig. 1. Action research cycle with pre-study (adapted from [35]).](/figures/fig-1-action-research-cycle-with-pre-study-adapted-from-35-2k6yqhdp.webp)
