Rehabilitation Engineering and  Design Research - Theory and Method Bodil Jönsson and Peter Anderberg

All scientific research requires problem definition and discipline, as well as concept and method formation. However, from a human point of view the most important problems are probably those found in areas where it may be difficult to provide such ideal research conditions. Rehabilitation engineering and design research is one such field. Nonetheless, we believe that these theory and method formation problems are manageable and our objective in writing this report is to demonstrate how.

The starting point for rehabilitation engineering and design research are human needs/wishes/dreams and its most important yardstick is the enjoyment and benefit it brings to users. The process thus begins with the individual and ends with the individual. At the same time, the method, and to some extent the language, of rehabilitation engineering research is that of technology – the technical solutions and their design demonstrate how problems have been interpreted and how technical and educational possibilities can be implemented. The field has many points in common with medical research, although rehabilitation engineering and design research concerns exterior tools while in medical research the tools are interior ones. But the goal is the same as that of medical research: to cure, alleviate and/or comfort.

Introduction

Rehabilitation engineering and design is a field at the intersection between technology, the natural sciences, the humanities, the social sciences, and medicine – fields that differ a great deal from an epistemological point of view. This report is an attempt to add to the theoretical and methodological basis of the field, to indicate where it may find theoretical support, to put forward some fresh ideas, and to create a usable theoretical foundation. Our thoughts are based on a document from 1997, Certec’s Core [1], which constituted a first attempt at summarizing and describing the theoretical and methodological basis for Certec’s work. The present document includes some parts taken directly from Certec’s Core, other parts have been expanded on, while some parts are new.

Products, processes, and knowledge

Rehabilitation engineering and design research comprises three areas:

1. The products with special emphasis on the initial design phase and subsequent design reflections when they are in use.
2. The processes (both the design process and the user process)
3. The gradual development of knowledge about users, processes, and products.

The field also includes those organizational contexts that encompass the products, the processes, and the relevant knowledge. It would be impossible to exclude knowledge of HMI and process awareness – it would lead to a rather pointless focus on gadgets. However, from an organizational point of view, the care services have laid the foundations for such a division by usually making a clear distinction between medical rehabilitation, habilitation, and assistive technology.

The central design concept

Internationally, the term design stands for usability, useworthiness, and esthetics - a wider application than in Sweden, where the main focus is on form and beauty. In this wider sense, it is also the task of the designer to listen to how the situation "talks back" (Donald Schön, [2]), not only during the development of a prototype but also during the entire life and reuse cycle of the product.

This requires much more extensive user research than is presently the norm. The Swedish Transport and Communications Board (KFB) is in the process of defining a platform for user research (1999). See also the anthology "Users in Action" [3], in which the chapter "Users in Action Learning" pp.91-107 deals specifically with rehabilitation engineering and design.

A Classical Perspective

Aristotle made the following division into three categories:

• Episteme
• Techne
• Phronesis

A brief look at these words tells us that two of the areas are still alive and well today: episteme (epistemology, the theory of knowledge) and techne (technology).

"Phronesis", however, is a word for which we have no active equivalent. Phronesis is about values and reality, about people and their actions. In the last decade we have seen renewed interest in "phronesis", particularly in the social sciences. Sometimes it is also used in design contexts. See, for example, Pelle Ehn’s preface to Löwgren & Stolteman Design av informationsteknik – materialet utan egenskaper, 1998 [4].

Phronesis is needed, Martha Nussbaum [5] writes, as "an attack on the claim that all valuable things are commensurable; an argument for the priority of particular judgements to universals; and a defense of the emotions and the imagination as essential to rational choice".

Phronesis is thus not scientific in the "episteme" sense, since epistemology mainly deals with scientific knowledge that is universal, constant in time and space, context-independent, and based entirely on analytic rationality. The knowledge relativism that is an integral part of phronesis is thus almost unforgivable in an epistemological approach. However, the connection between "techne", the reflecting knowledge of concrete action, and phronesis is better developed.

In an interesting section on methodology in her doctoral thesis "From green image to green practice" 1999 [6] Minna Gillberg writes "A phronetic research approach should focus on practice, because human action cannot be understood or judged through generalisations, static concepts or universal theories only, but rather to be found in the practical knowledge which builds on human experience. Therefore we must study practice, the concrete particulars of reality (the priority of the particular) that are complex and constantly changing."

Regardless of whether research is carried on in the domain of epistemology, technology/design, or phronesis, the only research result worthy of the name is new knowledge. Accordingly, in a research project it is seldom the entire process or the project result as such that is the actual research result; rather, it is solely the new knowledge generated in the project. It is thus essential to identify and define what this new knowledge consists of and to relate it to the existing knowledge – quite a delicate task, especially when the research is carried out partly in the domain of phronesis.

The Official Swedish Government Report 1999:4, God sed i forskningen ("Ethical Research Practice") p. 67, [7] includes the following statement concerning this interest in "new knowledge": "Professor Johan Cullberg uses a criterion for good science which fits in well with Popper’s solution, the so-called surprise criterion. This criterion means that it should not be possible to predict the result of an experiment or a study. There must always be an element of surprise if the knowledge that is generated really is new."

Epistemology – if possible

If something can be described by fixed concepts, generalizations and universal theories, a great deal is gained by doing so. Large areas of "Rehabilitation engineering and Design" can be dealt with within the framework of epistemology and can thus pride itself upon:

• its ability to systematize and accumulate
• its ability to articulate new questions
• its openness, transparency even, in its handling of methods and data
• its capacity to generalize on the basis of experience gained
• its openness to other perspectives which may make the results look different

In all these instances, epistemology strives for universality, context-independence, and non-relativism. This is advantageous – provided that it is possible and relevant. If not, the priority of the particular [5], [6] must apply, i.e. we enter the domain of phronesis. Here, the ideal is the Aristotelian agent, who is characterized as follows by Minna Gillberg: "The Aristotelian agent is a person whom we trust to describe a complex situation with full concreteness of detail and emotional shading, missing nothing of practical relevance." [6].

Why phronesis now?

We feel that it is somewhat double-edged to bring to light an area that has been dormant since the days of Aristotle. We have decided to do so because, firstly, we feel that the concept is needed and, secondly, we are now being helped along by the social sciences. We agree with Minna Gillberg’s statement in [6]: "In Martha Nussbaum´s writings about practical reasoning, which is her understanding of Aristotle’s concept of phronesis, I found a discussion that gave room for a deliberation that included imagination, empathy, perceptiveness and responsiveness. The discourse of phronesis (practical reasoning or wisdom) is very broad and complex. What I found so powerful and imaginative about MN´s writings was how she applied, in an interdisciplinary way, the Aristotelian tradition on our contemporary society and gave vivid life and profound meaning to the concept by combining ancient and modern literature."

Heuristics, Analysis and Synthesis

Heuristic reasoning is a fundamental element of rehabilitation engineering and design research since interpretations of a situation and its possibilities are rarely strict and final but rather multifaceted and provisional. It is necessary to be aware of this, especially in the initial phase (prior to more definitive solutions) and when evaluating the outcome of the solution one has arrived at. The situation must be allowed to "talk back" [2]. See also below under "Design".

Heuristic elements exist in most types of problem solving, including mathematical problem solving [8], [9], where one must

1. understand the problem – interpret the influencing factors, their mutual relationship, how to structure one’s reasoning,
2. find the connections between the unknown and what we already know, initially perhaps by simply exploring possible plans for finding these connections,
3. implement the plan, and
4. examine the solution – is the result correct and reasonable; could it have been achieved in some other way; can it also be used for other purposes?

From a heuristic point of view, the following are special features of rehabilitation engineering and design research:

• the scientist and the user are seldom the same person, and it is the user’s assessment of the outcome that is the most important end result (although not the only one – as mentioned above, the research result is the new knowledge that has been generated).
• the number of variables is so great that the choice of analysis level will determine how the problem is solved (see below under "More is Different").
• usually, the research subject is the prospective rather than the existing situation

In this connection, we would also like to point out that a person’s understanding of a system and its parts may be founded on intellectual one-sidedness: the analysis may be effective, but not the synthesis. The main point of the article "More is Different" by P.W. Anderson in Science, 1972, [10] , a key article reaching far beyond the field of science, is that reductionism – analyzing complex things into simpler constituents – has been found to be a successful scientific method and will most likely remain so. However, not even in the domain of natural science is it possible successfully to proceed in the opposite direction, that is, to constructivistically work one’s way up based on results at lower levels. Given knowledge at a basic level does not in itself construct relevant complex connections at a superior level.

Rather, on each new level genuinely new characteristics appear, which must be studied based on their own complexities. Anderson’s article highlights how this applies in the field of science and only touches briefly on the difficulties of the same nature that one encounters when working more closely with individuals – for example, the fact that there are more levels between ethology and DNA than between DNA and quantum electrodynamics.

In other words, the relationship between a system and its parts is intellectually unidirectional. Analyzing and breaking up a large problem into smaller parts can be a successful method in a complex situation, but operative synthesis based on partial results is seldom productive.

To sum up: problems and possibilities should be defined in their complete human and environmental context. In order for a relevant analysis to take place, one must then both break down the problem into lower levels (e.g. technically manageable levels) and constantly return to the original context when assessing the relevance and success of the results. And perhaps most importantly: a scientist in the field of rehabilitation engineering and design must form an image of the prospective reality on the basis of the present in that part of reality which is to be changed (including all its dependent variables).

Time Geography

Even with a very clear focus on context, the phenomena which are to be studied by breaking a whole into smaller parts sometimes lose their distinctive character. Essential support for avoiding such superficiality of thought can be found in the field of time geography. The creator of time geography, Torsten Hägerstrand, describes it as "mental diagrams which sort out our thinking and enable us to know the place of empirical studies in the overall picture" (from Tingens vidd, [11]). Time geography is thus a kind of "situational science", i.e. it is interested in situations in the true sense of the word. This observation is closely related to the considerations concerning case studies which are discussed below. In addition to the specifically geographical aspects – space and presence – case studies also involve personal, cultural, and social aspects. Thus, to a greater extent than in the field of time geography, case studies in rehabilitation engineering and design are dependent on being capable of including "phronesis" as well as "episteme" and "techne".

By looking at a situation from a time geography perspective one becomes more aware of the limitations inherent in extracting things and events from their context in order to subsequently analyze them by way of, for example, various experiments. According to Hägerstrand, it is easy to miss what he calls "vicinity" relationships, i.e. the proximity and simultaneity of different elements. "Vicinity" relationships comprise everything in the temporal and spatial world that interacts with the phenomenon in question.

"Vicinity" relationships are important in rehabilitation engineering research – in order for a technical device to be at all useworthy (this concept is defined in a doctoral thesis by Håkan Eftring, [12]) it must interact in an adequate manner with everything and everyone in the vicinity. If not, either the device or the vicinity must change. Technology, which can compensate for a functional limitation, assuage, or entertain (cf. the objectives of medicine: to cure, alleviate, and comfort) supplements or substitutes for assistance from other people. Sometimes technology empowers the user in a way that no other person can but, even so, the technology must fit into the user’s environment.

Hägerstrand uses Karl Popper’s concept of three different worlds:

1. the physical material world (natural science phenomena from the Big Bang onwards, as well as artifacts, i.e. phenomena created by man);
2. the world of ideas (thoughts, ideas, and interpretations); and
3. the cultural world with its rules, laws, and agreements.

In our case, this division is not sufficient – we would like to draw a razor-sharp distinction between the natural sciences and the science of artifacts. That which has been created by man not only influences the world of ideas and the cultural world but is also itself influenced by these two worlds – quite the opposite of the indifference to man of the Big Bang and the matter and laws it generated. Another basic difference between natural science and technology is the fact that natural science has a mechanistic structure while the structure of technology is teleological (see below).

Case Studies

Both the novice and the experienced researcher are regularly engaged in case studies – studying individual cases in depth, defining problems and solutions, and analyzing effects. Certec’s largest project, the Isaac project, [13], has now (in 1999) lasted more than 5 years. The project includes case studies of the technology development and design phases, [14], as well as of individuals and individual effects (Maria Sporre, logopedic study of improved communication [15].

In What Isaac Taught Us [14] we focus attention on a number of more general statements:

1. You cannot rely on needs to be formulated just because they exist. Technology can be a very good (and sometimes the only) means of showing that a problem exists and what it consists of.
2. You cannot know until you have tried.
3. The researcher is often controlled by his own mental pictures to the extent that the end result turns out to be amazingly like his original picture. Consequently, one should attempt to make this mental picture visible, for example in the form of a mock-up, so that others can see it, too, and criticize it as early as possible.
4. When an idea from the researcher’s high-tech environment is transformed into a product in the user’s no-tech environment (without even passing through a low-tech one) and finds its place there, almost anything can happen. But you need a "computer facilitator" and someone to brainstorm with. Continuously.
5. If the solution leads to the empowerment of the user, it can be a useworthy and powerful tool for change. If not, the effort will have been in vain.
6. Technical development is always much faster than one can imagine. And then it turns out to be completely different from what one had expected!
7. There is a close link between empowerment and the power of initiative. At the Pictorium day center, the multiplicity of digital pictures started a hitherto invisible inner process that was essentially different from that which the instructional pictogram teaching method had been capable of achieving.
8. A language used in alternative communication should preferably also be effective, and in use, outside its specific environment.
9. The importance of pictures when it comes to accessing the past cannot be overemphasized.
10. An assistive device (in this case digital pictures) can have a profound pedagogical impact on the users, but they may still not manage without it. The same assistive technology can thus be an educational as well as a compensatory tool. Users must have constant access to these tools, and holidays and vacation periods can lead to serious problems unless the device is also used in the home.

We agree with Martha Nussbaum’s [5] and the Danish social scientist Bent Flyvbjerg’s view that case studies are extremely underestimated as a research strategy. Interested readers are referred to Flyvbjerg’s book Rationalitet og magt [16] and above all to the brilliant "Chapter 8. Exemplets magt" ("The Power of Examples").

From Flyvbjerg, we have borrowed the following outline of how an information-maximizing selection should be constituted:

1. Look for extreme/aberrant cases

That which is extreme or aberrant can act as an eye-opener – a phenomenon is usually not exclusive to the aberrant individual, but she may help us see it. For example, the in-depth interview with Eva Gerdén in [12] makes visible a person who to some extent accepts her disability, but who definitely does not accept one of its consequences: constant dependence on other people. There is one aspect of the fact that "more is different" that is particularly noticeable for some people with functional impairments: the individual tasks that they need assistance with become something completely different when they are added together into a long chain. Having someone help you pick up a sheet of paper, or get a book from a shelf is very different from asking for help five, or ten, or fifty times. Even if the task of a personal assistant is to assist, the fact remains that they are human assistants and that ordering a person to do something is quite different from using an assistive device. It is not just a matter of showing consideration for the assistant – it is also about the fact that it is necessary to express instructions in words to a human assistant instead just carrying out the task independently with the aid of the technical assistant. Eva Gerdén has a very strong preference for technology (for example in the form of a robotic arm) over personal assistance – and she is very good at expressing it. _ThThis makes her a valuable extreme case.

2. Look for maximally divergent cases

It may be advantageous to select 3-4 cases that are maximally divergent with respect to a certain parameter – age, for example.

3. Look for critical cases

These are the cases where it is possible to say: "if it applies to him, then it applies to everyone". Or: "if it does not apply to him, it does not apply to anyone."

(A classic example to help the thinking process: In physics, when one wanted to show that a kilo of matter would fall just as fast regardless of what it consisted of, it was enough to do one experiment with one kilo of lead and one kilo of feathers. The densities of lead and feathers are so different that if these materials fall to the ground in the same amount time, the same will be true for all other materials. When the extremes have been covered, there is absolutely no reason to repeat the experiment.)

One Certec example of a critical case is the "IT and Learning" course which was offered in 1998 with priority for students with disabilities. With the aid of a variety of adaptations the majority of the students were able to complete the course. Furthermore, learning conditions for people with minor dyslexia have improved considerably through the use of technology and methods that allow for very slow reading and writing. In fact, in some cases everyone can benefit from solutions for people with disabilities.

4. Look for paradigmatic cases

This is what Flyvbjerg calls cases which can found a school for the field and serve as a metaphor for it. A scientific activity is recognized and acknowledged as good science on the basis of similarities with one or more examples of good scientific work. A paradigmatic case is precisely the type of case which founds a school for the field.

According to Thomas Kuhn, a specialist field without many good case studies is a specialist field without a systematic creation of models and, consequently, a weak specialist field. Certec wants to contribute to making the field of rehabilitation engineering and design a strong specialist field, and the best way of doing this is probably to carry out more case studies which can become models. We are particularly pleased that case studies are eminently suitable as a meeting-place for people with disabilities, researchers, and other interested parties - a meeting-place where the general and the particular can coexist and both deserve to be given prominence.

Technology, Design, Action Research, and Educational Methodology

This refreshing, strongly worded statement in God sed i forskningen ("Ethical Research Practice") (SOU 1999:4), [17], indicates that a research effort should not be required to apologize for having the aim of improving the human condition. On the contrary: the Commission wishes to emphasize that this is the obligation of every researcher.

In the social sciences this type of research, whose aim it is to bring about a change during the actual research process, is called action research. It goes back to 1947 and a paper by Kurt Lewin entitled "Action Research on Minority Problems" [18]. Despite the fact that action research has existed for more than 50 years it has not become firmly established. Sometimes it has been severely criticized because the same researcher observes, influences, documents, and analyses a change process that he is himself involved in. This is, of course, very different from the type of research whose sole focus is on describing and analyzing, and which has as a hallmark that the process should be outside the sphere of influence of the researcher. (This discussion is similar to the discussion concerning episteme versus phronesis, that is, it dates back over 2000 years rather than just a few decades.)

With respect to technical research there is usually no need to state the obvious fact that it is aimed at bringing about change. Neither do we know of anyone who has accused technological research of the weaknesses usually attributed to action research. One reason for this may be blindness to the inherent potential of technology to produce change. Technical research is, of course, to some extent clinically pure since the research subject is almost exclusively the technology itself rather than the user of the technology. The branch of research, user research, which may be initiated by the Swedish Transport and Communications Board would thus be extremely valuable.

Let us discuss in more detail the fact that technology is teleological, i.e. it has an in-built "in order to" basis (unlike the natural sciences’ mechanistic "because of"). For example, clocks are made in order to indicate time and telephones are made in order to make it possible to speak to someone at a distance. In all technical development work there is at least a vague notion that the technology in question should meet some need – it may be a need in one or many individuals or an improvement or an addition to existing technology which is already used by people (e.g. a "time-saving" technology).

Rehabilitation engineering research is even more distinctly teleological than other technical research. In general technical research the focus is not as explicitly on human needs as it is in the field of rehabilitation engineering research. Ideally, rehabilitation engineering research should begin and end in the person concerned in her own environment.

There is an epistemological connection between technology and educational methodology. Both are teleological. Both require stringent structuring of reality in order to make it more understandable and manageable. Although exposing structures may be just as essential to educational activities as it is to technical activities, it is much easier to have a tacit structure in teaching than it is in, for example, programming where the reasoning process is made explicit. Software consists of condensed thoughts.

If an educator is able to make explicit her educational diagnoses as well as the connections between diagnoses and measures taken, education can at least to some extent assume the structure of software. The work of one of Certec’s doctoral students is aimed in this direction (licentiate thesis by Eve Mandre, 1999, [19]).

The primary function of the software design process is to force the creator, the original educator, to make visible her reasoning as to what can and should be diagnosed and how diagnoses and pedagogical measures are related. Certec has had similar experience of how the use of rule-based relationships can be a way of extracting knowledge from experts through Charlotte Magnusson’s work on expert systems [20]. The expert in question, the person whose knowledge is to form the basis of the system, is forced to become more aware of his own thinking than ever before. Subsequently, once the software (for example the educational software) has been created it enables other people to perform a diagnosis according to the input structure requested and to obtain suggestions for possible educational measures. This will inspire substantive pedagogical discussions of an unprecedented kind – we will now have access to the other person’s internal relationship perceptions, i.e. her perceptions of how diagnoses and measures are connected. (The term "internal relationship perceptions", which is much more expressive than the traditional term "theories", was used by Torsten Hägersten in a letter to Bodil Jönsson in 1997.)

HMI

Rehabilitation engineering research must also adopt a position on the branch of research called HMI – the study of the interface between the user and the technology. HMI research is exceedingly important to Certec, but to a greater extent than in general HMI research we must take into consideration the social context in which the technology is used.

HMI has its origins in experimental psychology. Terms such as usability and adaptation were first used in 1979/80. There were attempts to use so-called iterative prototyping, but this type of work was often difficult to reconcile with requirement specifications. Usability design (with measurable usability) began in 1985. It was criticized, however, for measuring only that which could easily be measured, and in 1987 it was necessary to reinvent participatory design, this time as "contextual design".

HMI research has thus developed in the direction of increasing flexibility, which makes it difficult to define and describe. What exactly do HMI studies of the interaction between the individual and the technology have in common and in what way do they differ from other completely or partially overlapping fields such as ergonomics, human factors, and human/computer interaction?

Certec is mainly interested in how HMI for people with disabilities can benefit from general HMI research, both in theory and in practice. There are, of course, a large number of principles for the design of usable technology. Good technology means that the right parts are visible and that they convey the right message. For example, a button should convey the message "press me", a lever should say "pull me", and a handle "turn me". To make a device easy to use, the user must receive confirmation that he is doing the right things. What if a seat belt did not click when fastened correctly? What if a camera did not click when a picture has been taken? Other ways for a technical device to confirm that it has received information can be lamps that light up. HMI can speak to the need for, and the function of, these kinds of features.

A technical device should help the user remember and provide cognitive support. Preferably, it should be self-explanatory. The user should not be required to apply his energies to operating the device. It should be designed in such a way that he can concentrate on the task at hand, both physically and mentally. With respect to the physical aspect, the airborne frictionless mouse in Breidegard & Jönsson A Minimeter for Emma [21] is an excellent example of how a minimal amount of remaining mobility can be utilized. Donald Norman’s [22] statement that people want to learn how to handle tasks, not how to handle technical devices, sums up the mental perspective. In terms of design it is thus essential to bring the tool to the task, rather than the task to the tool.

The user should be able to create a mental picture of how the tool works. There should be a comprehensible relationship between the functions of the device and the controls used to operate it. Moreover, the device should be sturdy. Preferably, it should be impossible to make a mistake, and should the user still manage to make a mistake, the consequences should not be disastrous. Examples of this are electrical plugs that only fit in the right kind of outlet and cars that will not allow you to lock the keys inside them.

HMI and disabilities

The fact that technical devices must be adapted to the individual, and not the other way round, is particularly obvious where people with disabilities are concerned. They are not able to compensate for the deficiencies of the technology (the way other people often can) and thereby achieve a functioning system despite the lack of functionality. In other words, the interaction between the technology and the disabled person makes it necessary to tackle the design of technical devices so that they will be usable to everyone – not just 18-30 year old male technology freaks.

In high-tech airplanes, nuclear power plants, or the processing industry, the technology dominates its environment. The technology defines what should be done – a particular task is to be carried out and the device was manufactured to carry out that specific task. It is a different situation when it comes to everyday technology and assistive devices. In such cases, there is always an environment to consider and there are people around the user who can influence how well the human/machine system will work. In addition, to a high degree the user himself defines what is to be done. The focus is on the user and his/her needs, wishes, and dreams. There is a big difference between developing a personal robot to meet the needs of a person with a functional impairment and a factory robot for industrial use [13].

The importance of the environment makes it difficult, if not impossible, to test assistive technology in laboratories. In order for the finished system to be usable, the environment must be included in the development process. The natural way of achieving this is to, as far as possible, let the development take place in the environment where the device is designed to be used. In this regard, the development process at the Pictorium day center, as described in, for example, What Isaac Taught Us [14] is very close to ideal. At the Pictorium it is now difficult to distinguish between ideas coming from the researchers and those coming from the users.

This iterative process requires high-quality interaction between engineers/developers and users, and each prototype in an iterative process must also serve as a means of communication between the engineer and the user. The knowledge generated in the development process can sometimes be transferred to other technologies and other users. A particularly effective way of transferring such knowledge is through the technology itself. Compare also [24], licentiate thesis by Arne Svensk (to be published in 2000).

Design

Internationally, the term "design" comprises the integrated interaction between needs, ideas, visualization, form, environment (work environment, exterior environment and environmental issues related to the use of the product in the home, at work, or in the community), financing, planning, manufacturing, packaging, transportation, and communication between the prospective user and all parties concerned. Ideally, the term should also include the study of the final use of the product in the home, at work, or in the community, taking reuse and recycling into consideration in the actual development process. Design thus concerns both the design process – an iterative and partly non-sequential interplay between the individual and the artifact – and its results and their impact on the individual during the use phase. This means that rehabilitation engineering research has many points in common with design research and that it is a part of the branch of design research that we might call "design for useworthiness". As mentioned above, the concept of "useworthiness" is introduced and analyzed in [12].

Design research requires a kind of reflective practice. _I It is closely connected to case studies and time geography (as well as to episteme, techne, and phronesis) in that it is restricted to the situation in question. People, things and events are not taken from their context to be studied separately, rather the particular is often given priority: action-oriented and situation-dependent perceptions of relationships and of how they can be handled.

Donald Schön’s The Reflective Practitioner from 1983 [2] is an important milestone in the development of design as a science. In the book, he analyzes structures and limitations of reflection-in-action. The limitations are set by the fact that the complexity that we are capable of handling subconsciously paralyzes us if we try to make it conscious. Moreover, it is not always explicitly possible to teach lessons on something that is context-dependent (unlike matters that are context-independent). However, it is possible to learn in the relevant context. A phenomenon can thus be learnable without being teachable.

Design problems are rarely defined by the surrounding world, and there are few easy opportunities to use requirement specifications as a starting-point. Instead, the problem description must often be constructed from the situation itself, preferably using technology as a language.

Many people find technology fascinating and will gladly spend a lot of time on it. In addition, technology has the capacity to make things visible. This makes it a helpful tool when a user and those around her want to become aware of needs, wishes, and dreams, provided one is always deeply respectful of the individual. The following are some of the possibilities of technology:

1. Technology is honest

People with disabilities and those around them can interact with technology instead of waiting passively for someone else to take the initiative. Technology is what it purports to be, no more no less.

2. Technology is neutral

It is difficult, maybe even impossible, for a human being to be completely neutral. It is particularly important to recognize this when people are acting as representatives of others who have, for example, communication difficulties or other cognitive problems. In this type of situation, the neutrality of technology may provide greater respect for the integrity of the individual than a fellow individual, however well meaning, is able to do.

3. Technology is concrete

With regard to people with cognitive disabilities in particular, it is essential to have an object to show and to refer to when one is trying to understand needs, wishes, and dreams. Abstract questions or thoughts simply do not work at all.

4. Technology provokes

A thought, a conversation, or a lecture has something fleeting about it. A technical device, however, endures and in the end forces a decision. Sometimes technology brings a previously hidden contradiction to light, which must be dealt with.

5. Technology surprises

Even the most open of research approaches has limits to how far it is aiming. When trying to use technology as a language, we have often been surprised by the way reality has answered questions completely different from the ones we asked when introducing the technology.

6. Technology liberates

There are moments when neither the person with a disability, nor the people around him, notice that he is facing obstacles. If one has not seen them, it is obviously not possible to work towards removing them. Sometimes the people around him do in fact notice the obstacle but take it for granted that there is no technical solution. As a result, one puts a lot of effort into making living with the problem bearable instead. In a third case, the disabled person himself experiences the need while those around him do not see it. If the disabled person cannot express his needs, nobody else will do it for him. In such a situation, the use of technology can be a liberating educational method. The absurdity of the situation is revealed. We can discover the unknown by twisting around what is already known.

7. Technology empowers

In our experience, over the whole range from simple push-button commands to advanced computer technology, only the technology that leads to the empowerment of a person with a disability has any profound effect. Empowerment, however, brings revolution.

This line of argument is strongly supported by Schön [2] who argues that a researcher-designer, a reflective practitioner, lets the products and the situation talk back. One must try to solve the problem that has been defined but at the same time keep one’s eyes open for phenomena that are not included in the problem statement. New hypotheses can be formed from these phenomena. The generative metaphors gradually acquired in each new situation are invaluable to the designer. These can be called "it is like..." Schön calls it "seeing as", while Kuhn calls it "thinking from exemplars". Being aware of the "it-is-like" method means that there is no need to mystify, for example, intuition or creativity.

In 1973, Rittel and Webber observed that design often involves so-called wicked problems - problems that cannot be fully described or analyzed before being solved. They are, however, described and analyzed in the course of the design process. Compare this with our first lesson of ten from the Isaac project: needs will not necessarily be expressed just because they exist. "Technology can be a very good (and sometimes the only) means of showing that a problem exists and what it consists of."

Like Donald Schön’s, Donald A. Norman’s position in the field of design is indisputable. He has argued forcefully that people are analog, not digital; and that people are biological, not mechanical [26]. We have already mentioned the fact that people want to learn how to handle tasks, not how to handle technical devices, and that in terms of design it is essential to bring the tool to the task, not the task to the tool. However, this is so important that we wish to repeat it here.

Donald A. Norman has also provided the clearest specification of a number of user requirements for IT agents. One of the most important requirements is that the user must trust the agents. This is perhaps even more essential in rehabilitation engineering research – an autonomous technical device is only valuable to people with a functional impairment if they trust it.

Further discussions concerning design applied to or specific to the field of IT can be found in Löwgren, Stolteman, Design av informationsteknik [4].

Methods

In connection with functional impairments one often talks of "special needs". But people with disabilities do not really have special needs – their basic needs are the same as everybody else’s. What is special is the fact that the functional impairments may require special methods, technology, or theories to help ensure that needs, wishes, and dreams can be fulfilled despite the functional impairment. This calls for special technology, special education, special care, and special imagination.

The methods used in rehabilitation engineering research should be subordinate to the imperative ethical requirement of rehabilitation engineering and design research, which is that the starting-point should always be the wishes and views of the individual. Usually, the person with a disability is often both the "research subject" and a "collaborator" in the design process.

We find support for this reasoning in three basic documents:

1. First, Article 27 of the United Nations’ Universal Declaration of Human Rights:

2. Second, Rule 13 of the United Nations Standard Rules on the Equalization of Opportunities for Persons with Disabilities:

"Rule 13. Information and Research. States assume the ultimate responsibility for the collection and dissemination of information on the living conditions of persons with disabilities and promote comprehensive research on all aspects, including obstacles that affect the lives of persons with disabilities." [28]

3. Finally, "Ethical Research Practice", Official Swedish Government Report 1999:4_:

/.. The whole research process as well as the role of the researcher must be subjected to a strict ethical analysis, primarily by the researcher himself. This ethical analysis requires a good knowledge of the facts and an awareness of the values that may be gained or risk being lost. When important values are at stake it is morally responsible to say yes as well as no to research. Moral responsibility also means taking responsibility for not carrying out an experiment." [29]

Ethical considerations are always close to the surface in rehabilitation engineering and design research (Fält and Jönsson, What is Right? [30]). The Internet may bring such considerations to a head. For this reason, Certec has elaborated a special Web Code of Conduct: http://www.english.certec.lth.se/code.html .

Tools for thought and action

There is no obvious boundary between design theories and design methods. Methods are tools for thought and action. Successful rehabilitation engineering and design research requires a good thought and action toolbox and the ability to select the right tools. An important step may be to divide the methods into different compartments.

There are many ways of providing the field of rehabilitation engineering research with methods and a language which make it possible to connect and distinguish between its different branches. One uniting aspect is that it may be appropriate to question, at the very outset, whether the solution should imitate fully the solution for a non-disabled person (the parrot method), have the same purpose but a different form (the chameleon method), or be completely different and only retain its fundamental characteristics, its very core (the poodle method). Whichever method is used, it is self-evident that, as far as possible, one should avoid designing special special technology. The chances of success in the use of the device are much greater if it is based on existing technology and if, for example, computer applications are compatible with standard software.

1. The Parrot Method

If it is possible to imitate, like a parrot, the way a non-disabled person would handle a certain situation, this may be the best solution (at least from a social perspective). This means that the system consisting of the person with a disability and her technology is capable of doing exactly what she would otherwise have been able to do without technology and she chooses exactly the same approach to problems which other people around her can handle without the aid of technology. For example: glasses, prostheses, corrective medication, wheelchairs, etc.

This is the most common approach in the field of rehabilitation engineering. The parrot method can be successful, and sometimes this is where one must begin. But it is important to let the situation talk back and to follow up to ensure that one has solved the right problem. One can sometimes avoid ending up in a dead end by attempting to define the function one is aiming for before beginning to try to solve the technical problem. "Wouldn’t it be better to try the chameleon method?" "Or maybe even the poodle method?"

2. The Chameleon Method

In the chameleon method the aim is to perform the same task as the non-disabled person is able to do, but it is not meaningful to try to imitate the way in which it is carried out.

Instead, like a chameleon, one tries to change the color of the solution by changing technologies to achieve the same result.

The task of mailing a letter is a good example. For people who use a wheelchair this poses so many problems that it is hardly meaningful to keep practicing the task, to develop assistive technology for it, etc. In this case, it is better to refrain from mailing letters (atoms) and just send the information (bits) via electronic mail instead. There may be a need for a customized user interface, i.e. special assistive technology may be required to enable the person to use a computer. In that case, this should be the focus of rehabilitation engineering, not trying to copy other people’s old letter routines in a parrot-like manner.

Examples of other chameleon solutions for people with visual impairment, for example, include using Braille, speech synthesis or audio books instead of ordinary text (the purpose is the same as it is for sighted people: being able to take in something that has been documented). Guide dogs are another example (the purpose is the same as for sighted people: being able to move about independently).

3. The Poodle Method

Like Goethe’s metaphor in Faust, this is about getting to the heart of the matter; about finding the innermost part of the dream, the wish, or the need. Even with technology, it may not always be possible to do what one wants to do. And sometimes it may be possible, but it might not be worth it to make the exact original dream come true at any cost: maybe it is just the feeling one wants to achieve.

In that case, one must look deeper. Was it a desire for a challenge, intellectual or physical, that was the driving force? Would it be possible to find an activity that can be physically experienced just as much or even more, an activity that will make the body buzz with exhaustion and joy? Perhaps there is an altogether different activity that could provide the same intellectual challenge? These are the types of activities that should be supported by rehabilitation engineering, not the original one which, in fact, can no longer be achieved.

To sum up: As was pointed out earlier, one advantage of this type of structure is that it can bring out common as well as distinguishing aspects of different types of rehabilitation engineering. One distinguishing factor is that a physical disability that has been compensated for need not affect cognition at all, while visual and hearing impairments often lead to different ways of using sensory impressions and sometimes also to different cognition. A congenital neurological disability or a developmental disability, acquired brain damage or a mental illness may result in very special ways of perceiving and interpreting the world, which others may be unable to understand and empathize with.

The parrot, chameleon, and poodle analysis can be applied to the distinguishing level as well. One example is the prevailing view among neurotypical people (an adjective I learned from a woman with autism) that neuro-atypical people are non-empathetic. Perhaps this is nothing but a parrot projection: a matter of viewing the actions of the neuro-atypical person as failed attempts at mimicking the neurotypical person. Perhaps the heart of the matter is that the autistic person interprets the situation in a completely different way, but then acts logically (and empathetically) based on her interpretation. Perhaps the least empathetic of the two is the neurotypical person who sees the situation only from her own point of view and measures the other person’s actions by her own yardstick?

1. Jönsson, B. (January 1997). Certec’s Core. [Online] URL http://www.certec.lth.se/doc/certecscore/index.html
2. Schön, D. (1983). The Reflective Practitioner. Hampshire: Ashgate.
3. Jönsson, B., Svensk, A., (February 1999 ). Users in action learning: The explosion of digital pictures and its importance to the power of initiative, internal relationship perceptions, and the language of differently abled persons. In Karlsson, M., Östlund, B., KFB-Meddelande 1999:8 :Users In Action (pp 91-107). ISBN 91-88371-21-2.
4. Löwgren,J., & Stolteman, E. (1998). Design av informationsteknik – materialet utan egenskaper. Studentlitteratur, Sweden.
5. Nussbaum, M. (1990) Love´s knowledge, Essays on Philosophy and Literature. Oxford University Press.
6. Gillberg, M. (1999). From green image to green practice. Lund Studies in Sociology of Law.
7. (SOU 1994:4. (1999). God sed i forskningen, 67. [Online] URL http://utbildning.regeringen.se/propositionermm/sou/index.htm)
8. Polya, G. (1962). Mathematical Discovery. NY:Wiley&Sons.
9. Polya, G. (1990). How to solve it. NY: Penguin Books.
10. Anderson, P.W. (1972). More is different. Science vol 177, number 4047, 393-396.
11. Hägerstrand, T. (1991). Om tingens vidd och tingens ordning. Texter av Torsten Hägerstrand. Stockholm.
12. Eftring, H. (1999). The Useworthiness of Robots for People with Physical Disabilities.
13. Isaac Project Overview [Online]. URL http://www.english.certec.lth.se/isaac/index.html)
14. Jönsson, B., Philipson, L., Svensk, A. (1998). "What Isaac Taught US" ISSN 1101-9956 [Online]. URL http://www.certec.lth.se/doc/whatisaac/index.html
15. Maria Sporre (2000). Digitala bilders kommunikativa funktion för människor med talhandikapp.
16. Flyvbjerg, B. (1991) Rationalitet og makt. Akademisk Forlag, ISBN 87-500-3011-6. [Online]. URL http://www.i4.auc.dk/flyvbjerg/default.htm
17. (SOU 1994:4. (1999). God sed i forskningen, pp 23. [WWW-page] URL http://utbildning.regeringen.se/propositionermm/sou/index.htm)
18. Lewin, K. (1946). Action Research and Minority Problems. Journal of Social Issues 2. 34-46.
19. Mandre, E. (1999) Från observation till specialpedagogisk design.
20. Magnusson, C. (1998) Expert Systems and Learning
21. Breidegard, B., Jönsson, B. (1999). A Minimeter for Emma.
22. Norman, D.A. (1988). The psychology of everyday things. N Y: Basic Books.
23. Norman, D.A. (1993). Things that make us smart. Reading, MA :Addison-Wesley.
24. Svensk, A. (1999) Licenciate thesis, fall 2000.
25. Rittel, H., Webber, M.. (1973). Dilemmas in a general theory of planning. Policy Sciences 4:155-169.
26. Norman, D.A. (1988). The Invisible Computer. Cambridge, MA: The MIT Press.
27. (UNIVERSAL DECLARATION OF HUMAN RIGHTS, Article 27. [Online] URL http://www.un.org/Overview/rights.html
28. From the United Nations’ Standard Rules. [Online] URL http://www.independentliving.org/STANDARDRULES/StandardRules1.html
29. SOU 1999:4. (1999). God sed i forskningen, 23. [Online] URL http://utbildning.regeringen.se/propositionermm/sou/index.htm

    30. Fält, A., Jönsson, B. (1999). What is Right? Everyday legal and ethical questions encountered by Certec. [Online] URL http://www.certec.lth.se/doc/whatisright/