|
program (02) = areas of study |
Design science studies go beyond the basic linking of existing disciplinary models, to the expected functional and utility consequences of design decisions, as well as to the unintended consequences on individuals, society, and the environment. By studying these consequences, you can contribute to the advancement of existing models and the production of new models more relevant to the process of designing artifacts. You can also study the process of design, using past experience to advance the management of design processes, as well as our knowledge of the process and structure of innovations.
Design science program studies can be organized into the following major research domains of inquiry: decision making processes and preference structures; decision management; structure of innovation and social models; economic, marketing, business, and policy models; ergonomic, human variability, and aesthetic models; information, complexity and systems optimization; design context of high technology . You can study these research domains of inquiry in specific contexts, for example, sustainable and life-cycle design, aesthetics, design of highly customized products, designing for an aging population, and design and government policies.
The research domains and application contexts reflect the interests of the program faculty and students and are likely to evolve over time.
Decision Making Processes and Preference Structures
Major advances have been made in the social sciences by using simple rational models as a benchmark against which to compare actual behavior by humans (e.g., the Nobel prize-winning prospect theory developed by Kahneman and Tversky that challenged expected utility as a descriptive model of decision making). These findings suggest that there may be systematic deviations from rational decision making. The design process provides a new arena against which to test basic principles about decision making, in a setting that is more realistic than the typical laboratory of the social scientist.
The optimization framework also has a set of qualitative tools that can facilitate the understanding of a decision problem. These qualitative tools can be used to develop new tests of models in the social sciences, which in turn can feed back into the design process (e.g., prospect theory can replace expected utility in an optimization model).
This use-focused approach also encourages a careful examination of the process a person follows when making a choice. What information do individuals consider when making a choice? What features are considered? Do individuals focus on the same attributes that are relevant to designers? How much choice should the individual have? Can too much choice paralyze decision making? Does the presentation of tailored artifacts, where a design model selects a subset of artifacts to present to the individual, lead to better decisions? Do such related practices (e.g., Amazon's practice of recommending books, Dell's practice of allowing a user to configure their own computer) lead to better decisions, and, if so, under what conditions?
Decision management is a term that describes the various things that managers do, wittingly and otherwise, that affect how and how well the people in their charge make their decisions. The product design process is rife with decision problems. Clearly, the ultimate success of a product or a design group rests heavily on the adequacy of the myriad design decisions made every day. Particularly in the case of complex products, the design process is a collective rather than an individual enterprise, such that supervision of the process must consider the collaborative as well as the individual activities of each group member. It is easy to cite countless instances where decision management practices have significantly shaped the fortunes of virtually every kind of organization. So what is the nature of decision management customs in design groups?
The specific research questions to be addressed can be conveniently framed within the context of a particular conceptualization of decision management and decision processes known as the 'cardinal issue perspective'.The so-called 'possibilities issue,' for example, considers the range of potential consequences of a particular design decision, rather than the more commonly studied questions of how likely it is that already-recognized consequences will actually occur. Any leader of a design organization would have to exercise decision management skills that somehow succeed in getting designers to decide routinely in ways that today are probably exceptionally rare.
Generalizing from decision making in other domains, we can speculate about the character of decision making and decision management in design groups. But how valid are these generalizations? Early research will rely on observational methods pioneered in anthropology, which have been used with some success in other fields, e.g., marketing and medicine. For instance, research team members will become participant observers in design groups. They would use standard ethnographic methods but also ones that are informed by the categories of decision scholarship more generally. Among other things, the resulting observations may yield a variety of specific propositions regarding the relationship between product success and how well the possibilities issue is addressed within different design groups.
The Structure of Innovation and Social Models
Examples of bad product design are easy to collect: We find them when we attempt to move about during our everyday lives. Product designers have a tremendous influence on how easily, and how well, individuals' goals are met each day. However, most efforts to understand the cognitive processes in design focus on evaluation. Focus groups, prototypes, user testing, and design critique sessions are often successful in identifying problems with existing design. But beyond characterizing examples as 'good' and 'bad' for the human users, psychology has relatively little to say about how to create innovative designs.
Innovation, as opposed to invention, requires a structure that leads to good designs. Is there a way to describe and evaluate models of the design process? If yes, then such models will lead to specific preferred structures around which we can build analysis models and organize our tasks. Methods from cognitive ethnography can be used to study this question of structure. How do designers incorporate knowledge of the consumers' needs, available technology and materials, and economic constraints, to design a truly 'human' product? Norman has characterized good design as 'user-centered,' where the human interacting with a designed artifact finds it very useful and easy to use, and readily adopts it in favor of existing alternatives. But while good product design seems to have consensual, even testable standards for evaluation, little is known about the cognitive processes leading to innovation in design.
Within psychology, work on innovation has focused on creativity and insightful problem solving. Studies of creativity have identified some aspects of successful processes. However, most psychological studies have focused on isolated problems designed to require no background knowledge, and have used undergraduate students rather than design experts as subjects. These studies fail to make use of settings where real innovation in design takes place.
An observational study of expert product designers known for their innovation, intuitive feel for the user, and product success can reveal the cognitive and social processes present in good design. Cognitive ethnography studies across products and teams can suggest processes that contribute to successful innovation.
Previous research on complex cognitive processes, such as pilot performance in the cockpit, has demonstrated that studying teams may facilitate uncovering the cognition involved because the team members frequently express their thinking through interactions with others. In addition, the physical environment allowed interaction with devices and displays, so that physical gestures and changes of location gave further clues about the influence of information sources in cognition. Thus it was possible to identify 'critical incidents' that characterize key moments in the process where outcomes are determined. Similar studies can lead to a deeper understanding of how to design for interactivity as well as the critical incidences in the creative design process.
Economic, Marketing, Business, and Policy Models
Marketing models have been developed over the last 50 years to analyze consumer choice and many other aspects of product design. Underlying these models is a presumption that the engineering process can be decoupled from measuring consumers' needs and subsequently communicating product benefits to them. The prevailing overarching methodology, in fact, is one which gauges the success of a product or group of products based on three disjoint tasks: (1) measuring consumer reaction to products profiled in terms of their underlying 'attributes', (2) choosing promising target products so that designers and engineers can attempt to build them; and (3) taking the resulting candidates, those faring well by (1) and (2), and forecasting their market potential at a specific point in time. Little is known about the effects and artifacts of decoupling these procedures, and even less is known about how to fully integrate them or allow them to intercommunicate over time.
The most successful marketing methodology –used in the design of thousands of products and services– is conjoint analysis. Yet conjoint, and all other methodologies, treats products as groups of disembodied attributes, such as durability, speed, price, etc. How those attributes are chosen in the first place, how they interact, whether they are the same for everyone, what levels (e.g., different price points) are relevant to consumers, how engineers and aestheticians figure into the process, all are considered peripheral in the determination of 'optimal' products. A chief research goal would be to restore the many interwoven strands of the design process to an equal footing in formal optimization methodologies.
Marketing theorists and practitioners focus great energy on the assessment of 'new' products. A typical 'new' product might take an existing product and add additional features to it, or combine features from products already available. Less typically, it might be something truly novel, offering features or user benefits unavailable in any form. Real-world products tend to skew towards the former, 'line extension' sort, simply because they are easier to assess, communicate and predict about. Truly novel products, particularly these for which design and 'emergent' uses play a large role, tend to wither in the long and grueling march to market introduction. Among the intended outcomes of design science research is how to avoid stifling true innovation simply because it is easier to put numbers to the mildly and predictably new, rather than the transformative. In particular, we must study how to allow key elements of design and aesthetics to enter marketers' vocabulary and decision processes.
Numerous practical design issues can be addressed. Marketers at present do a good job figuring out what to produce to fulfill consumers' wishes in two situations: When the product is 'high involvement' (a lot of thought goes into choosing it) or is relatively simple (it has few dimensions of importance to consumers). Roughly speaking, our methods work when we know a lot about what consumers are looking for, and when there isn't much they are looking for. The vast majority of products and services lie outside these demarcations: anything for which there is a large experiential component; where aesthetics matter (art, furnishings, appliances); where there are long-term, complex considerations (homes, major durables); where there is a great deal of future uncertainty and investment (educational programs); where important attributes differ greatly among users (autos); where the playing field is rapidly shifting (anything high-tech). In each of these situations, there is a non-trivial psychological, aesthetic or engineering-based component. As a result, the best methods used to measure, quantify and plan the product creation process are difficult to apply or unreliable. In such cases, marketers fall back on methods rife with potential for bias: asking people what they want, and letting managers wade through their verbal replies. It is in this comparatively uncharted realm of product realization that design science research break new ground.
A key idea pervading product design in marketing is heterogeneity, where consumers differ in what they seek from a particular product. In recent years, demography has introduced two powerful differentiation variables: life stage and globalization. The emergence of older consumers as an economic force is poised to redefine multiple industries over the coming decades. Among the challenges of designing and marketing products will be to account for the differing needs of large segments of consumers at various stages of their lives, particularly as they progressively require assistance. Product designers will also need to become increasingly conversant with cross-cultural designs: those that can compete in the global marketplace without sacrificing usability in any particular nation or culture. Merely extending or adapting existing designs, those fashioned for the 'typical' consumer –Western, young, educated, media-literate– will not adequately address the needs of such diverse groups.
Ergonomic, Human Variability, and Aesthetic Models
Current education in human factors and design has little in common. Design instruction consists of pointing students toward handbooks of prescriptive guidelines. In addition to a need for more and better models of human capabilities and requirements, there is an even greater need for research in methods for integrating knowledge of human requirements into the design process and for teaching students about these requirements.
An important source of variance in the performance and success of products designed for use by people is the people themselves. Design quality and value are often affected more by the variance in the human users than by the variance attributable to the product's hardware. Consequently, optimization of products used by people will benefit from consideration of human variance through robust design methodologies.
Human factors are represented in the optimization problem by models of anthropometric variability, postural variability, objective performance criteria, and subjective responses. Anthropometric variability describes the user population with respect to body dimensions; postural variability (e.g., a seating position) is partly predictable from body dimensions, but considerable residual variance remains that must be taken into account in the design evaluation; objective performance criteria include safety-related measures, such as proximity to dangers or fields of view (e.g., 95% of drivers must be able to see a point on the ground two meters in front of the vehicle); subjective criteria include the cost of 'dis-accommodation,' the situation where when a person's preferred component location(s) or posture are not accommodated by the design, a concept tightly linked to marketing evaluations.
A domain of direct interest and application is the design of medical equipment. For example, design of implants and prostheses is often done without sufficient consideration for the procedure itself. The required design of the necessary instrumentation follows the design of the prostheses resulting in increased complexity, cost and potential for errors. A holistic approach, as envisioned here, can have dramatic impact on medical health care costs and the comfort of patients and doctors.
Human factors research on interface and visual displays has traditionally defined interface effectiveness with criteria such as legibility or difficulty of target search and information access. The consideration of aesthetics is often ignored in human factors analysis of interfaces and displays. Artists and designers, in contrast, have long treated aesthetics as a major aspect of their work; however, they rarely employ experimental methods to examine the validity of their hypothesis and they mainly describe aesthetic terms in qualitative or subjective languages that do not easily allow for quantitative evaluation. Previous research on computational modeling and experimental investigation of interface and display aesthetics has resulted in preliminary quantification of the effects of three compositional elements: balance, symmetry, and grouping. Similarly, aesthetic evaluation of a wide range of interfaces including automobile and webpage design has been documented. Much more research is needed to include more design variables (such as color, shape, curvature, density, motion) and to integrate these models with other computational ones.
There is also a rich history of work in quantifiable aesthetics, such as golden section proportions and its generalizations, by classical and modern masters in the fields of architecture and industrial design. Further developing and linking aesthetic models with engineering functionality models, an effort popularized in Japan as Kansei Engineering, is an important area of design science research.
Information, Complexity, and Systems Optimization
From an engineering perspective, design of artifacts has been successfully modeled in a mathematical design optimization framework. Design optimization involves a mathematical statement of design objectives to be optimized (minimized or maximized) as functions of the design variables. Design restrictions are represented by equality or inequality constraint functions of the design variables that must be satisfied by an optimized design solution.
Design and product development in modern organizations can be modeled as distributed multilevel mathematical optimization frameworks. Complexity introduced through the interdisciplinary nature of design science can be addressed in this context, at least partially. The assembly of disciplinary models and methods for matching their requirements is an important research question. There is also an underlying need for an information technology infrastructure that can support the data collection and management necessary for building the models, as well as the communication among various decision-making models.
Frameworks for complex system optimization traditionally have been static, involving the definition of a system hierarchy, development of disciplinary/subsystem models, and mapping of their interactions. If the hierarchy changes the modeling process is repeated. However, modern design environment hierarchies are not static. Rigorous strategies for system optimization that maintain integrity of information flow in an evolving hierarchy is a major research challenge.
There is considerable uncertainty in the mathematical design optimization models: computed responses of the physical system are based on imprecise environmental parameters, designs cannot be realized exactly, cash flows, product demand and human behavior can be forecasted only in a statistical sense. Therefore the adopted design strategies must be formulated to deal with such uncertainties.
The Design Context of High Technology
The thinking and methods of design science can be studied and implemented within the context of specific technologies or application domains. For example, design science methods can be applied to analyze the impact of new technology results in the areas of electric propulsion, energy, smart materials and structures, medical equipment, sustainable design, and designing for an aging population. Students can undertake studies in these domains based on their own interests and the interests of design science faculty.
