The ubiquitousness of failures in new information systems designs and implementations indicates that the design methodologies that the field has developed are inadequate or faulty. This paper points to the body of knowledge developed from design research as a potential source of expertise for developing design methods that are more successful. The paper describes a meta-theoretical method for classifying theories in the Information Systems field, and identifying areas that need further development.
There is widespread criticism that the designing and implementation of Information Systems (IS) solutions are frequently problematic in terms of late completion and being over-budget (see, for example, The Word, 2000, Lyytinen et al., 1998, Standish Group, 2001, Gladwyne Software Surety Inc, 2000). A primary purpose of Information Systems research is to facilitate the designing of information systems. These criticisms imply that the ways that information systems are designed are not sufficiently well developed: that the design methodologies devised on the basis of IS research are inadequate or faulty. These problematic approaches to designing information systems contrast with the more successful approaches used for designing other forms of artefact, many of which are considerably more complex and have greater capitalization than IS solutions.
This paper contributes to rectifying the above situation by describing theoretical approaches that have emerged in the field of Design Research: the main discipline that focuses directly on improving designing and designed outcomes across different domains. The paper draws attention to the close relationship that existed some 40 years ago between IS research and Design Research, sets out some key definitions, and maps out the different research and theory-making themes that have emerged as important in design research. The paper describes a method of meta-theoretical analysis to identify areas of weakness in the balance of theories in the literature of IS research, and concludes with suggestions for improving designed IS outcomes.
In the 1960s, a beneficial symbiotic relationship existed between the fields of IS and Design Research (see, for example, Simon, 1969, Jones and Thornley, 1964). Since then, the field of IS has developed its own design methodologies (see, for example, Adams et al., 1985, Astrom and Wittenmark, 1990, Bielawski and Boyle, 1997, Curtis, 1995, Long, 1982, Shneiderman, 1998), and the previously close theoretical relationship between the fields has declined, although IS theories have been used in design research (see, for example, Court, 1995, Coyne et al., 1990, Dasgupta, 1991, Eder, 1989).
Terminology relating to designing is problematic (see, for example, Hubka and Eder, 1988, Love, 2000c, Eder, 1966, O’Doherty, 1964, Liddament, 1996). Definitions used in this paper are based on the epistemological analyses of Love (2000b, 2001 (accepted)-b, 2001 (accepted)-a):
· ‘Design’ - a noun referring to a specification for making an artefact or for undertaking an activity. A design is the basis for and precursor to the making of an artefact (including systems, theories, methods). Creating an artefact does not necessarily depend on a design, for example, in Art and Craft)
· ‘Designing’ - non-routine human internal activity leading to the production of a design.
· ‘Designer’ - someone who is, has been, or will be designing, i.e. creates designs
· 'Design process' - any process or activity that includes one or more human acts of designing in association with one or more other sorts of activities.
· ‘Design Research’ – research and theory making about designing and designs.
Other meanings and uses of ‘design’ are avoided, especially the epistemologically problematic use of ‘design’ as an entity with agency, in, for example, ‘”design” creates better solutions’. The above definitions separate the activity of designing from other human activities, and identify acts of designing in many situations in which they have been overlooked.
IS Research has two main foci: the activities of designing information systems (in some cases with their associated business case and plans), and analysis and prediction of the physical and theoretical characteristics of information systems. These equate in design research terms to ‘research into designing’ and ‘research into designs for different forms of artifact’. Like the design processes of most products, successfully designing information systems goes beyond addressing purely technical issues. It relates to all of the three main themes of design research:
· Human aspects of designing - Research into designing as a human internal activity (including all human stakeholders).
· Methodological aspects of designing - Research focusing on design processes and developing of methodologies to improve design processes. This includes research into how the activity of designing co-joins with many other activities like calculating, programming, drawing, analysing data, collecting data etc
· Technical aspects of designing - Research into problems, designs for solutions, and the physical and theoretical relationships between them. This includes information about or from stakeholders.
Research into human aspects of designing - understanding and modelling designers’ internal activities is important. Designing is a primary aspect of individual human functioning, underpinning how humans create change and respond to problems. This aspect of design research lies at the heart of understanding individual and group behaviour and motivation. It is relevant to understanding and reducing IS failures because it offers the greatest insights into including human social, environmental and ethical issues into designing solutions (Love, 1998b).
IS systems are increasingly likely to draw on research in this realm because of the increased use of affective mechanisms in information systems such as customer relationship management systems. Affective issues are involved in several other aspects of designing IS solutions including:
Traditional information systems for management focus on business process. The recent shift towards e-enabled customer-centric business processes indicates a need for a fuller understanding of customers. Design research can also offer some insights in this area because customers themselves are involved in designing, especially in the ways they use the products and services of others as design support tools to ‘design’ their lives. These design support tools include, e.g. using an online loan calculator, and using on-line catalogues and advice to design a new lifestyle. The way that customers design their lives points to a role for design research to helping IS professionals and researchers involved in building information systems to support advertising, marketing, corporate planning and organisational planning professionals who can be reconceptualized as managers to ‘customers as designers’. This approach offers potential competitive benefits that complement traditional customer research methods.
Research focusing on design methodologies has been a major theme in design research over the last 50 years or so, particularly in the Design Methods movement from which the UK-based Design Research Society and the US-based Design Methods Group originated. Hundreds of design methodologies and models of ‘design processes’ that are potentially relevant to the design of information systems have been devised to efficiently and effectively address design problems across a very large number of sub-disciplines. These methodologies focus on different issues. Some focus on bringing together the necessary elements of a design for an artefact. Others are aimed at the ‘public face’, or the interface, of a designed artifact. In an IS context, the first sort of design methodologies might involve data types (name, address, company, transaction, date etc). In the second, the focus of the methodologies might be the answers to questions such as: “Who needs to know the data?’ ‘Why?’, “ How will the system supply them with that data?’, ‘ How are the user’s feelings affected by how the data is presented?’, ‘Which parts of the systems are easier / more difficult to access?, ‘What is the most natural flow of events and tasks within the system?’, ‘What alternative pathways are in the system?’, ‘Why, and how are pathways signaled to the user?’.
Other types of design methodology outcomes from design research that are relevant to praxis in the development, implementation and management of information systems are:
In these latter methodologies, information is repackaged in forms that are ‘designer-friendly’. Examples of such packages in a software environment are: high level languages, RAD environments and software toolboxes e.g. Delphi, C#, and Cognos, These are language and information system packages that support program designers to function at higher conceptual levels that are more effective and efficient. The underlying design methodologies consist of processes to develop these collations of programmatic and informatic elements into easily accessible and carefully structured products to support designers in creating solutions without them needing to reprocess primary knowledge structures and data.
Research into technical aspects of designing focuses on problems, designs of solutions, and the relationships between them. Research in this area follows these three themes:
· Research focusing on problems: for example, formal ways of defining problems to help designers produce solutions to those problems effectively and efficiently.
· Research focusing on designed solutions: for example, identifying the ways that humans interact with particular designed solutions, and how the characteristics of particular designed artefacts align or comport with the characteristics or behaviours of other designed artefacts.
· Research focusing on the relationships between problems and designs of solutions to those problems. The aim is to automate problem-solution scenarios to move them out of the arena of designing and into routine processes. Design researchers have successfully undertaken this task across a variety of disciplines, perhaps most visibly in the engineering disciplines where many tasks previously requiring high-level human skills in designing and judgement are now automated via computer-based information systems.
The complex nature of human circumstances means that there is rarely a single right or correct solution to creating a design to resolve a particular problem: in most cases designing involves minimising the adverse compromises necessary to create a solution that is whole and functional. The findings and theories of design research offer support for IS professionals to more effectively use other relevant, bodies of knowledge, especially relating to the cognitive and informatic processes that are a part of designing.
So far, the focus has been on the topoi regarded as important in design research that may contribute to improving IS design methodologies and designed outcomes. Another way of exploring the interrelationships between the fields is to focus on the structure and dynamics of their theories (Stegmüller, 1976).
Different fields contribute theory to human understanding in different ways, in different areas, at different levels of abstraction, and with different amounts of coherency and justification. A structural meta-theoretical approach to analyzing theory focuses on categorizing theories and concepts in terms of their relationships at different levels of abstraction. The most abstract levels relate to ontology and epistemology, and the least abstract relate to individual humans’ direct interactions with their environment and selves. Between, with different forms of hierarchical dependence to each other, are the day-to-day theories used by professionals and researchers.
Several meta-theoretical hierarchies have been devised to map out different categories and relationships between theories. These include: Giddens’ (1987) subdivision of Social Theory into large, small and middle-range theories; Reich’s (1994) three level model of theory in Engineering Design and Artificial Intelligence; Franz’ (1994) four level structure for theories about designing in Architecture; Popper’s (1976) three incommensurate ‘worlds’, Ullman’s (1992) complex taxonomies of mechanical design theory, and Howard and Lazear’s (1990) seven different areas of understanding. Each of these has been devised for a particular purpose.
The nine level meta-theoretical abstraction hierarchy described below in Table 1 was developed to assist in clarifying the relationships between different fields of study associated with human agency (such as designing IS products), and for checking the coherency and sufficiency of general theories, theoretical perspectives and sub-fields of study (Love, 1998a, Love, 2000c, Love, 2000a, Galle, 2001). This meta-theoretical hierarchy locates theory in a human context in which designing (the skills at addressing wicked problems and creating new knowledge and artefacts) is included as a core aspect of human existence. The hierarchy consists of a generic structure that is modified for analyzing theory in different situations (see, for example, Love, 1998b, Love, 2000c).
Table 1: Meta-theoretical hierarchy of concepts and theories
Level |
Classification |
Description |
|
1 |
Ontological issues |
The ontological basis for theory making: human values and fundamental assumptions of researchers, designers and other theoretical foundations and axiomatic entities. |
|
2 |
Epistemological issues |
The critical study of the nature, grounds, limits and criteria for validity of knowledge. |
|
3 |
General theories |
Theories that seek to describe human activities and their relationship to designed objects and human environments. |
|
4 |
Theories about human internal processes and collaboration |
Theories about the reasoning and cognising of individuals involved in designing and researching, of collaboration in teams, and of socio-cultural effects on individuals’ behaviours. |
|
5 |
Theories about the structure of processes |
Theories about the underlying structure of processes of designing and researching based on domain, culture, artefact type and other similar attributes and circumstances. |
|
6 |
Design and research methods |
Theories about and proposals for methods and techniques of designing and researching |
|
7 |
Theories about mechanisms of choice |
Theories about the ways that choices are made by designers and researchers between different elements, designed objects, processes, systems or other types of possibility. |
|
8 |
Theories about the behaviour of elements |
Theories about the behaviour of elements that may be incorporated into designed objects, processes and systems |
|
9 |
Initial conception and labelling of reality |
This is the level at which humans descriptions of objects, processes and systems are coined, e.g. ‘a vacuum cleaner’, ‘a car body’, ‘a groyne’, a ‘database’, ‘sitting’ at a ‘desk’, ‘hearing’ ‘noise’, ‘smelling’ ‘fumes’ from an ‘exhaust’ and ‘watching’ ‘sunsets’. |
The focus of this meta-theoretical hierarchy is on theory qua theory. Theoretical objects and elements are any things (not necessarily physically real or concrete) that are represented in theoretical terms. This is a position that aligns with many IS conceptualisations, e.g. object-oriented programming. The meta-theoretical hierarchy is reflexive and theories from one perspective may well be the objects or elements of theory for another and vice versa.
The balance of theories and concepts at different levels in the meta-theoretical hierarchy are different for different bodies of literature and different disciplines. Theories of IS can be mapped onto this meta- theoretical hierarchy in a fairly straightforward manner; made easier because of the organized nature of information systems. The diagram below demonstrates a hypothetical mapping of theories of Management Information Systems and Design Research across the hierarchy (The distribution of theories at each level is a personal estimate for demonstration only and is not based on any formal research).
An additional measure plotted in the above example is the sum of the contributions from the two fields.
The profile of this sum can offer increased understanding about the relationships between two fields. Where it is relatively constant in spite of the relatively uneven distribution of theories from each field it may indicate that the fields are complementary, and may have a symbiotic relationship in terms of research, theory building and solution creation.
Three axioms relating to the meta-theoretical hierarchy are:
These axioms in conjunction with a meta-theoretical mapping of the distribution of theories for a field such as Information Systems offers an overview of the coherency of the theoretical developments in the field, and can indicate areas in which there is a shortfall or neglect of theory development.
In many technical fields, the application of the above method frequently indicates weakness in relation to levels 1, 2, 4 and 9. Compared to other technical disciplines, IS has a relatively strong theoretical presence at level 9 due to research on interface designing and other ergonomic issues. Disciplines close to the physical sciences are likely to be relatively light in levels 1 and 2 because the ontology and epistemology of physics (on which many scientific disciplines are based) is axiomatically defined at these levels. This does not apply to Information Systems because the development of IS solutions also requires addressing human subjective issues that involve human values, and hence the field of IS would be expected to possess a rich and detailed theoretical foundation in levels 1 and 2.
Theories about human internal functioning (level 4) are epistemologically important to ensuring theoretical coherency in many disciplines to which they may at first seem relatively irrelevant. Any discipline whose knowledge is intended to support the development of artefacts for human use presumes and depends on theory about human cognition, feelings and behaviour. For other theories to be coherent and fully justifiable in a discipline, this implies it is necessary to make explicit the tacit level 4 theories.
Design methodologies are limited to the extent to which they can be successful in commercial use when the focus is only the technical aspects of problems and designs for solutions, and does not include human subjective issues based on values and ethical and aesthetical sensibilities (Dilnot, 1982). For design methodologies of IS (levels 5, 6 and 7), the above meta-theoretical hierarchy offers the means of checking the external and internal aspects of design methodology models. It provides the basis for identifying the assumptions on which a method depends, and the lower level theoretical constructs whose relationships it models. It also enables each methodology or method (as an abstraction) to understood in terms of its associated meta-theoretical pathway of associated theories on levels 1-9. These meta-theoretical ‘pathways’ in many cases can be envisioned in terms of metaphors. For example, theories that relate to the metaphor ‘humans behave like robots’ form a complete, identifiable and coherent set across all the levels. These ‘human robot’ theories form a different theory set from those related to for example, ‘humans behaving as creative geniuses’. With the meta-theoretical hierarchy, this metaphoric approach forms an additional tool for creating appropriate design processes and methodologies, and identifying the weaknesses in methodologies.
Information Systems activities of designing, understanding, managing and developing solutions all involve many acts, small and large, of designing that can be only partially described in terms of information and information flows.
In this paper, improving the design methodologies of Information Systems has been explored in terms of topic areas of design research, and through the use of a meta-theoretical hierarchy. The analyses suggest that the field of Information Systems might helpfully draw on findings, theories and theoretical perspectives of design research in relation to:
Many of the factors that supported the original symbiotic relationship between the two fields remain and this suggests that it may be helpful to explore the benefits available through joint research and for reviews of literature to include new developments in both disciplines.
Adams, D. R., Owles, V. A. and Powers, M. J. (1985) Computer information systems development : design and implementation, South-western Pub. Co., Cincinnati.
Astrom, K. J. and Wittenmark, B. (1990) Computer-controlled systems : theory and design, Prentice-Hall, Englewood Cliffs, N.J.
Bielawski, L. and Boyle, J. (1997) Electronic document management systems : a user centered approach for creating, distributing and managing online publications, Prentice Hall PTR, Upper Saddle River, N.J.
Court, A. (1995) The modelling and classification of information for engineering designers, University of Bath.
Coyne, R. D., Rosenman, M. A., Radford, Balachandran, M. and Gero, J. (1990) Knowledge based design systems, Addison-Wesley Publishing Company, Reading MA.
Curtis, G. (1995) Business information systems : analysis, design, and practice, Addison-Wesley, Wokingham, England ; Reading, Mass.
Dasgupta, S. (1991) Design Theory and Computer Science, Cambridge University Press, Cambridge.
Dilnot, C. (1982) 'Design as a socially significant activity: an introduction', Design Studies, 3, 139–146.
Eder, W. E. (1966) 'Definitions and Methodologies', In Gregory, S. A. (ed The Design Method, Butterworths, London, pp. 19–31.
Eder, W. E. (1989) In International Conference on Engineering Design 1989: proceedings of the IMechE, Vol. 2 Mechanical Engineering Publications Ltd Bury St Edmunds, UK, Harrogate UK, pp. 1307–1320.
Franz, J. M. (1994) 'A critical framework for methodological research in architecture', Design Studies, 15, 443-447.
Galle, P. (2001) 'Metatheory in Practice', Design Studies, 22, 101-102.
Giddens, A. (1987) Social Theory and Modern Sociology, Polity Press, Cambridge UK.
Gladwyne Software Surety Inc (2000), Vol. 2001 Gladwyne Software Surety Inc.
Hubka, V. and Eder, W. E. (1988) Theory of Technical Systems, Springer-Verlag, Berlin.
Jones, J. C. and Thornley, D. G. (Eds.) (1964) Conference on design methods, The Macmillan Company.
Lazear, D. (1990) Seven Ways of Knowing: Teaching for Multiple Intelligences, Hawker Brownlow Education, Australia.
Liddament, T. (1996) 'Metamorphosis of Design Vocabulary', Design Studies, 17, 303-318.
Long, L. E. (1982) Design and strategy for corporate information services : MIS long-range planning, Prentice-Hall, Englewood Cliffs, N.J.
Love, T. (1998a) Social, Environmental and Ethical Factors in Engineering Design Theory: a Post positivist Approach, Praxis Education, Perth, Western Australia.
Love, T. (1998b). Social, environmental and ethical factors in engineering design theory: a post positivist approach. Unpublished PhD thesis, Mechanical and Materials Engineering, University of Western Australia, Perth.
Love, T. (2000a) 'A Meta-Theoretical Basis for Design Theory', In Durling, D. and Friedman, K. (eds), Doctoral Education in Design, La Clusaz, France.
Love, T. (2000b) 'New roles for design education in university settings', In Swann, C. and Young, E. (eds), Re-inventing Design Education in the University, School of Design, Curtin University of Technology, Perth, pp. 249-255.
Love, T. (2000c) 'Philosophy of Design: a meta-theoretical structure for design theory', Design Studies, 21, 293-313.
Love, T. (2001 (accepted)-a) In Computational and Cognitive Models of Creative Design CognitionHeron Island, Queensland.
Love, T. (2001 (accepted)-b) 'Constructing a coherent cross-disciplinary body of theory about designing and designs: some philosophical issues', Design Studies.
Lyytinen, K., Mathiassen, L. and Ropponen, J. (1998) 'Attention Shaping and Software Risk: A Categorial Analysis of Four Classical Approaches', Information Systems Research, 9, 233-255.
O’Doherty, E. F. (1964) 'Psychological Aspects of the Creative Act', In Jones, J. C. and Thornley, D. G. (eds), Conference on design methods, Macmillan, New York, pp. 197–204.
Popper, K. (1976) Unended Quest, Open Court, Illinois.
Reich, Y. (1994) 'Layered models of research methodologies', Artificial Intelligence in Engineering Design and Manufacturing, 8, 263-274.
Shneiderman, B. (1998) Designing the user interface : strategies for effective human-computer interaction, Addison Wesley Longman, Reading, Mass.
Simon, H. A. (1969) The Sciences of the Artificial, MIT Press, Cambridge Mass.
Standish Group (2001), Vol. 2001 Standish Group.
Stegmüller, W. (1976) The Structure and Dynamics of Theories, Springer-Verlag, New York.
The Word (2000), Vol. 2001 The Word.
Ullman, D. G. (1992) 'A Taxonomy for Mechanical Design', Research in Engineering Design, 3, 179-189.