As a contrast to descriptive (or "disinterested") research which aims at finding out how things are (or how they have been), the target of normative study is to define how things should be.
Some researchers think that if you want to change things, you first have to find out how things are and then you can start planning the changes. Such a research process has, indeed, been used in many research and development projects, though it is arduous. It can include phases such as:
Another, simpler approach is often possible in the case - which is quite common - that the informants who assist at the initial registering of facts, can simultaneously evaluate the existing state of things and also participate in developing the proposal. Normative recording takes account of both facts and opinions concerning the object of study, and means also merging the first four operations in the process into one. The whole process thus shrinks to contain only three phases:
Of course, such a simplified process is possible only on the condition that the informants know well the object of study and are able to evaluate it meaningfully. This is the case when they have been selected from one or more of the interest groups defined for the project, in other words, when they are stakeholders in the project. Often it will be impossible to have all the interest groups present at the evaluation, and it remains the task of the researcher to collect the missing facts and opinions later. To facilitate this task, you can at least try to select the participants from the most important groups of interest.
Besides saving work, evaluative recording of facts can bring with it advantages such as:
Although the normative style of data recording is not always possible (perhaps because competent evaluators are not available at the moment of recording facts) some situations where it is possible, are especially the following:
Preparations before normative recording. Before proceeding to empirical operations, it is advisable to keep in mind that gathering of data is usually the most arduous and expensive phase in a research project and it deserves to be planned carefully. Remember, too, that it is in many ways related with operations that come before and after it in the research process. If there are weaknesses in the preceding operations, it heralds difficulties in the empirical work, and if data collecting is haphazard there will be problems in their later analysis.
The checklist below contains a few important things that you should have done before launching empirical operations in a large scale. If it, however, seems too difficult to make a decision on them right away, you can instead start by studying the problem first in a small scale before completing these tasks and starting to collect the main bulk of empirical material.
However, before starting to gather information the hard way empirically, it is useful to check whether somewhere already exists comparable information that you could apply to your purposes, at least as an example of approach in a normative project. Below are discussed two potential sources of such existing knowledge: publications and data banks of feedback from customers.
If your normative project has its origin in a frequent problem in a profession, it is possible that a similar problem has been encountered elsewhere and a normative project has already been carried out, perhaps successfully. If it has been documented in a book or in an article, you may find the document by means of a bibliographical search. Note, however, that when using earlier research findings the problem is not only finding them, but also assessing whether they are applicable to your present situation. On the page about bibliographical studies you will find a discussion about assessing the usability of the found information, from your point of view, which includes asking yourself questions such as:
|Is the information true?||Source criticism|
|Can the information be applied to your context: Is the object of study essentially similar? Are the findings statistically significant? Are the data recent enough?||Validity Assessment|
|Can we use the information for development purposes?||Utility Assessment|
Many companies that have stayed long in the market, have feedback arrangements which enable them to learn from the mistakes that sometimes happen. In this purpose, two approaches are often used:
In enterprises that sell personal service to the clients it is natural and easy to collect the clients' assessments on the service that they have received. As a contrast, in modern industry producer seldom has a direct customer relationship with the consumers of the product. Wholesalers and retailers now serve as intermediaries. The voice of the consumer reaches the growing and improving production plants with increasing difficulty if special channels are not created for it.
Probably the most common method to obtain customer feedback is complaint management. Besides treating the original problem of the customer by e.g. replacing the faulty product, the enterprise can collect the information content of the complaints from those sources where they have been directed:
The best category to investigate among these will be the complaints made to the company itself; it will also be easy to gather the complaints made to the retailers of the product. The other groups are seldom investigated.
Complaint management is, however, a crude method for gathering customer feedback because it is involved mostly with complaints and costs and lacks completely all positive feedback, ideas for developing the service and the quality of products, and ideas for better products.
There should be a parallel channel for not only complaints but for any comments that a customer might wish to give to the company. The customers have, of course, the option of sending a letter, but in practice this is seldom done. Instead, it might be worthwhile for a company to complement their existing www-pages with the option of transmitting customers' comments to the company. This can be done with HTML forms which can accept feedback either as fixed-choice alternatives or as opinions in free form, or both. In either case, the forms can be made so that the feedback arrives conveniently pigeonholed for each different product or service of the company so that they are easy to file and analyze.
Lately have appeared a growing number of WWW-sites for general feedback concerning any product that the public wishes to discuss, i.e. anybody who wants to say something about a new product can start a specific line of discussion for it. These www-pages are independent, i.e. they are not operated by the makers of the products which shall be discussed on the page. In the future, if the discussion becomes lively on these pages, it might be a good idea for a manufacturing company to study the responses now and then.
When analyzing the voluntary feedback that people send in as their opinions you should keep in mind that these people are far from being a random sample of all the customers (in fact, they are a "sample of volunteers") and therefore their opinions may be biased, i.e. differ from those of other customers who did not give feedback. Beside other possible types of bias, there is normally the phenomenon of the "absent middle": those people who have no decisive opinions on your product or service will send no feedback, which means that in the feedback that arrives two sorts of opinion predominate: the distinctly positive and the clearly negative ones.
If you wish to compare the amounts of complaints concerning the different products of the company, you should keep in mind that they are usually most frequent in the running-in phase of a new product when the novelty still suffers from "children's diseases". Another rise in complaints can be expected when the useful life of the product turns to its end and failures become more frequent. The "bathtub curve" on the right is from Abbott, 1989 p. 127.
Feedback from employees. Suggestion Scheme is an arrangement that has long existed at many industrial plants. It has mostly been used for gathering ideas from workers and employees for improving the production process, but it could assist in improving the product, as well, though this option has seldom been used. The system usually consists of the following parts:
Public critique, when it exists, could be a valuable source of feedback to a manufacturer, although it is, per definition, normative and thus the judgments depend on the values and goals of the critic or of his organization (see Point of view of normative study). Usual points of departure for public critique include:
User point of view. Consumer organizations and their journals, and lately a growing number of technological journals are continually testing cars, boats, computers, cameras, hifi and other new products. The purpose is to help the customer to select the right product for his needs, taking into account all the aspects that can be relevant to its owner and long-time user. These aspects can include all the usual goals of product design which normally are measured and assessed on prefabricated scales, weighted and summed up as value engineering tables so that the result as accurately as possible agrees with the average user's point of view. Tests can include both scrupulous laboratory testing and prolonged intense practical operation in order to uncover potential weaknesses in the product. Pertinence and reliability of such tests are usually good.
Similar tests for service activities are becoming more and more common, as well.
Views of colleagues. A competent colleague's critique of a service or of a proposed or executed design could often be invaluable, because he can profit from his experience of the requirements that similar service or product normally is expected to fulfil. He is thus often able to simulate the customer's preferences quite successfully in the evaluation.
However, the content of professional critique depends much on its purpose and address. When the profession is young, the professionals need to get a footing for it by convincing the customers on the usefulness of their services. As a contrast, when the profession is well established, the discussion usually turns more inwards, into professional topics, which can degrade its relevance for the users of its services and to other outsiders. This can happen especially at the time of a paradigm change of a comprehensive descriptive theory or of a normative design style, i.e. when the younger generation of masters is struggling for influence in the ranks of the metier. Such a situation is described quite acidly by Bourdieu (1994, 62):
The history of a field [of art or science] advances in the fights between those that have a name (as a writer, a philosopher, a scientist, etc.) and their challengers, as they are called in the sport of boxing. In these fights writers, schools and works of them get older. Those that already have created masterpieces (and have won a standing in the field) fight in order to keep it, to immortalize themselves as classics, and to stop all development. Their challengers in turn can make a masterpiece only by thrusting into the past these established predecessors. ... The climbing new avantgarde tries to regenerate the foundations of the genre. They insist on returning to the roots and to the original purity.
Good professional practice of design, i.e. the merits (if any) in the product's design itself, are understood differently depending on place and time, but often mentioned merits include the skill of handling contrasting requirements in a balanced way; clarity, precision and strength of the result, novelty and originality of the form, taking tradition in account skillfully, which tradition can be either classical or vernacular. What is "good" design, was already discussed earlier, in the paragraph Professional designer's view.
Generally designers do not often want to put on paper their opinions about the works of their colleagues. One reason probably is that it really can be difficult to analyze designs in writing, because a great part of the competence is learned from masters to pupils, the skill exists mostly in tacit form and cannot be fully clarified as explicit theory. Many concepts are used without clear definitions, and in fact an important share of collegial evaluations appears not as written criticism but by selecting (by a jury composed of eminent designers) produced works to a national exhibition or to a professional publication.
Views of exhibition visitors. It is quite a tradition of art criticism is that the critic just describes his impressions and reflections during the short time that he inspects each work. Critiques in newspapers on exhibitions of designed products often follow this "impressionistic" tradition. Such writings thus contain just one person's opinions, albeit that the writer usually is an expert in the field, and they normally contain only a few aspects of the object. Sometimes an object whose only merit is that it deviates from the common line in the exhibition gets good marks in such an assessment. Impressionistic critique is therefore only occasionally useful for product development purposes. However, this style of quick and superficial inspection of objects perhaps a little resembles the situation in a business shop, when a client is choosing a product for himself, which means that this type of criticism can sometimes help designing products intended for impetuous clients. The questions around immediate appeal of products are discussed further on the page Evaluating Normative Proposals.
Public criticism often examines a large number of products in parallel. Consequently it often fails to focus on the products' specific strengths and weaknesses themselves, and instead simply puts the question: which product is best in this exhibition? Answering such a question is much easier than laborious speculation on all the aspects of a prolonged use of a product, but the resulting critique will be much less useful for the creators of products.
Receiving passively complaints (see above) and other spontaneous feedback can give some useful hints to the company, but often their sporadic content does not give answers to exactly those specific questions that the company would want to ask, for example, concerning the reasons why the sales slow down or potential improvements to the product that technological development has recently made possible. Indeed, many companies that have long been selling the same product or service, have arrangements for active gathering of feedback where exactly the right questions can be put to the right people.
Active feedback collection does not much differ from typical survey research with questionnaires or interviews. Per definition, the population consists normally of the users of the product or service in question. If you are selling a service, it will be easy to present questions to your customers, but if you sell physical products you cannot always meet your customers nor know their names and addresses. Many companies have started building up a customer register by enclosing in the package of each sold product a prepaid return envelope in which the client can give his or her name and address. For products that are sold on the internet, customer registration with a HTML form is common and easy.
In the case that there is no customer file nor time to build up one, sometimes a sufficient number of product users can be registered when they are, for example, visiting the product after-sales service or a spare parts shop. Note that the samples that depend on the customer's willingness to co-operation are often a little biased, see Non-random Samples.
Once a contact to a customer has been established, the normal method is to give them a questionnaire with return envelope. A modern alternative (beside the internet form, discussed above) is an arrangement where the client can call, paying just a local call, to a number where a machine records the feedback. The client gives his response by following the instructions of the machine, either by pushing the buttons of the phone, or by dictating an oral message.
When planning the feedback questionnaire you should note that the evaluations in itself can be difficult to decipher without some additional information. A layman can estimate as 'excellent' the same product that an expert with higher expectations judges as 'poor'. Therefore you should try to register beside the opinion, also an indication of the person's level of expectations, cf. Expectation.
Note that both customer satisfaction and dissatisfaction consist of various components such as the price of the product or service, its usefulness in different situations, its symbolism and status, durability and maintenance. Some of these components may be unsuitable from the customer's point of view, while others may be perfectly in order. Therefore, it is seldom very useful to get a total utility value of the product - instead the customer should be allowed to differentiate his answer.
As a medium of reciprocal discussion questionnaires and structured interviews are insufficient in the sense that when it is the company that asks the questions and the customer who answers them, the company will get no new ideas. Therefore, the customer should be allowed and encouraged to put forward such ideas that nobody in the company has come to think of. In this respect, a thematic interview, a discussion over the phone, is better than a questionnaire because it allows the consumer to present his latent information and wishes while it also makes additional questions possible and allows the parties to discuss new ideas and develop them together. If the customer is really interested in finding improvements to the product, he/she could be asked to participate in brainstorming or testing of the next model of the product.
Note finally, that to ensure a sufficient answering percentage, you have to reward the respondents for their trouble by, for example, giving them useful information about the use or care of the product they have bought.
When selecting the method of normative empirical recording of the existing state of the object of study, the first question is, do you need to study empirically the factual activities or products that are the object of your study, or can you do without your own empirical observations and measurements and simply collect both the facts and evaluations from the participating evaluators?
For answering this question, you have to consider how intimately your informants know the thing they shall speak of. Can they reliably describe and competently evaluate it right away from the memory? Can they imagine improvements to the existing state of things? If yes, you can probably use the interrogative methods, described below.
If no, you need to arrange the research design so that it includes or at least simulates the factual object of study. Methods for this are described in Empirical Normative Recording, discussed later on.
An intermediate approach (not discussed separately in the following) could be that you arrange two (or more) meetings with your informants, and in the meantime these try out the proposals for improved action that you have collectively prepared at the first meeting.
When defining the group of people who you will ask to take part in the initial sessions of recording facts and evaluations, the selection is sometimes self-evident. For instance, when you want to develop the activity of a work group it is this group that you will mostly work with, and in the development of a new product, you will try to select the evaluators among the target customers, if these are known and available. Generally, you will prefer people from the most important interest groups defined for the project, in order to diminish the need of later modifications to the proposals.
The normal descriptive methods of interrogation are as such suitable for gathering facts as well as evaluations. If the required number of informants is small (i.e. the study is "intensive"), you will usually select interview as your method, because it allows you to present additional questions and thus you can get deeper into the reasons behind the respondents' preferences. Besides, if the respondent perhaps has some ideas about an optimal theoretical state of the object of study, an interview is the best method for putting them down on paper. Questionnaires usually give more superficial and less reliable data than interviews, but they can be more practical in so extensive studies with a large number of respondents, especially when these live widely apart.
Questionnaires. When recording evaluations and preferences, you will normally want to discuss each topic in two questions. The first of these, the "objective" one, deals with the factual state of things, and the second one takes care of the respondent's subjective evaluation of the state of things.
If you, for example, are gathering the experiences that your earlier customers have had with your products, the first question can specify the product model that the person has factually used, and the second question then asks about the evaluation, such as in the examples below.
1. Have you lately used a portable telephone? Which model was it?_______________
2. How would you assess its battery life?
Another example of a study that examines the need of improving the air conditioning:
1. If you know the highest temperature (measured in shade) in your workplace on warm days, please give it here: ______________
2. Does the high temperature in your workplace annoy you? Tick one of the boxes below.
Yes, it annoys me often
It annoys me occasionally
No, I do not mind it
Self-reporting or "experience diary" is a handy method for recording a person's activity, such as the use of a product, and her evaluation of this activity. For using the method, the researcher has to prepare a special kind of questionnaire where the respondents can report their own actions and also their evaluation of it and possible comments.
The questionnaire usually contains a page for each day, with printed questions that depend on the information that the researcher wants to get. If the study concerns how the day has been used or at which hour certain operations have been done, the page can be a subdivided in hours. As an alternative, if only certain type of occurrence is of interest, there can be empty boxes for reporting just these, with relevant questions to be answered. For example, when you want opinions on a car model's usability, you can design a set of questions to be answered always when coming back from a car ride.
To get more particulars from certain episodes, one possibility is to give the respondent not only a questionnaire but also a cheap no-refill camera and instructions on when to use it.
Self-reporting is an easy and cheap method, and it will eliminate the disturbance caused by the presence of the researcher. It has also the advantage that it can capture infrequent events in inaccessible places, like a use of a laptop computer during a journey.
Other examples of normative recording of evaluations are market research projects that are made to give a basis for new product development, and Active Feedback Collection arrangements for some large producers of cars, for example.
Interviews. For gathering evaluations and preferences from people, an interview is often the best method. When planning the interviews, you have to consider whether you want to gather information on personal or collective relations with the object of study. If personal, you will want to arrange a personal interview separately for each respondent.
A group interview, instead, might be more suitable when the object of study relates to ways of living, to work or recreation, or to prevailing attitudes, usages, values and meanings about them. All these are created while the individual lives in his or her social group, they do not effectually exist when the person is isolated from the group and in such an isolation they cannot be studied, either. As a disadvantage of group interview can be seen that in a large group usually some members talk most of the time and others are more subdued, though some researchers think that if this is the normal way of acting in a group, it is as well that it becomes recorded in the project of research, too.
If you wish that all participants have their voices heard, you can propose to the gathering that they agree on the manner of discussion as something like the "Rules of a Democratic Debate" sometimes applied in Swedish development projects. These state, for example, that each participant has an obligation not only to put forth his or her own ideas but also to help others to contribute their ideas.
Discussion in an existing work team. Sometimes it happens that a work group that already exists in a business or other organization, comes in contact with a researcher and asks for advice in a problem in their activity. Such an existing group is often an ideal source of descriptive facts and normative information, because these people know intimately the matter of study and often their main problem are too many opinions about improvements. For organizing the development of the activities of such a group, action research is often a suitable method.
Action research is a method in which the researcher temporarily joins the target community, and, with his theoretical tools, helps the community to solve the problems it is facing. The changes that are needed for correcting the problems are defined and accepted in a series of seminars where all the members of the group take part. The logical process - which is described in more detail on another page - usually consists of the following, repetitive cycle:
The modified style of action can then either be adopted for permanent use, or it can be taken as the starting point of a new, similar cycle of the project, until a proposal, acceptable for all, has been found.
Another type of a normative group interview - or a group innovation team - is the focus group. It is often used to give grounds for the development of a new product. There are usually from five to ten voluntary participants, selected among the known users of the product or from important future customers, who are invited to discuss the expectations for a future product. It is to be admitted, though, that it is usually impossible to find a representative sample of future customers or other potential interest groups to participate in a focus group, which lowers the reliability of the results.
The procedure in a focus group is strongly goal-oriented, which makes it possible to prepare in advance material for facilitating the work. These can include an opening speech for declaring the objectives and agenda of the meeting, samples of earlier models of the product, and preliminary descriptions of the future product and its use, such as pictures or mock-ups (see Method of presentation).
Focus groups often resemble ordinary club meetings with agenda, secretary, and a discussion leader which should try to encourage even the more reserved participants to express their opinions. For the event of discussion drying up the leader should have a set of prompts which, of course, must not be loaded, giving the impression that the leader is expecting a particular answer.
The discussion is usually recorded on tape or video which the researchers can use later for making a summary of noteworthy opinions. Afterwards, this summary can be discussed once more with either the original participants or in a new focus group.
The method of focus groups is explained more fully in Jordan p. 141.
All the interrogating methods, described above, allow gathering simultaneously information about the present state of things, about its acceptability and about possible or desirable directions of developing it. A drawback of the interrogating method is that the respondent's description about the state of things is perhaps not always so exact nor reliable than the researcher could wish. A remedy to this is that the facts are not asked from the evaluator but they are collected by the researcher simultaneously with the interview.
The above would mean using two research procedures in parallel: an empirical gathering of facts such as observation of activities or measurement of physical objects, and on the other hand gathering opinions of what was observed: how acceptable is the state of things?
Who are the persons that should carry out the mentioned two operations? The normal practice is that the researcher or his assistant does the objective measurements, but who is competent for the role of an evaluator? This is a question that should be answered on the basis of the targets of the project and the interest groups defined for the project. Sometimes the researcher himself or a special expert in the field could do the evaluation. Sometimes it could be the person that has assisted as an actor in the observed activity, especially if he or she belongs to an interest group mentioned above, though in this case you have to consider whether other interest groups should be present, too. In any case, you have to take into account the practical availability of persons from these groups.
In the following are described two possible combinations of empirical recording and immediate evaluation of an object of study. The details of the method, of course, depend on the nature of the object: activities and static objects require somewhat different methods of observation and recording.
Evaluative observation of an activity, i.e. observation which includes evaluation, is an efficient and quick method for discovering possibilities for improving an existing activity, on the condition that there are persons competent to evaluate these possibilities.
Between intensive and extensive observation there is no fundamental difference, but the practical arrangements differ often a little because of the different number of cases to be studied. Intensive study seeks to optimize a more or less unique case of activity, while extensive study aims at developing generally applicable improvements and the number of cases is accordingly larger.
Intensive observation, in other words an attempt to develop an existing activity but not its general theory, is often organized so that all the people that participate in the activity, also have the right to participate in the development project as legitimate members of the team. The approach is thus nearly identical to the normative interviews and work team discussions mentioned earlier, with the difference to them that here first the participants perform or simulate the activity in question and the researcher records it, after which follows a general discussion of possible improvements.
Another variant of intensive observation is carried out by the researcher, and the persons that are observed have then only the role of objects of study. This approach of Methods Engineering aims at improving the productivity, safety or other desirable characteristics in the activity of one or a few workers only.
Testing Prototypes can also be mentioned here because it often involves trying out and evaluating not only a product but also the activity of using it.
An example of many-sided normative observation is Sirkka-Liisa Keiski's project for developing a new type of kitchen furniture for the elderly. She observed, with the help of a video camera, old aged people in their home kitchens which she also measured. She noted that the kitchens were apparently designed for people with average motor abilities and in many respects impractical for aged people. The inhabitant demonstrated how he or she worked in the kitchen, and explained the disadvantages which Keiski then analyzed. Finally she constructed a mock-up kitchen and this was then tested during a second interview session. See a photo of the mock-up testing.
Extensive observation. Most methods of descriptive Systematic Observation can be used for normative purposes by adding a normative evaluation to the objective recording procedures.
The basis for evaluation will often be defined already in the initial normative targets of the project. If the object of normative study is, for example, industrial production, the target is often to improve a given characteristic in the activity, such as its economy, for example. Because many important characteristics of industrial production have already been much studied and there is much theory available about them, the normative evaluation of an industrial activity can often be done simply by the researcher on the basis of the already existing theory.
When studying products or other physical objects, intensive and extensive approaches are in principle similar, but in practice the number of objects will affect the choice of methods.
Intensive normative study aims at improving only one or a few objects, and the number of objects from which data will be recorded is also small - though it can be advisable to be on the safe side and study a few objects more than just the selfsame objects that you intend to improve, which can allow making fruitful comparisons between them. Because of the small number of objects, it is often possible to study them thoroughly in the genuine environment with all the relevant properties and relationships (i.e. the study is "holistic" or "idiographic"), thus achieving a deep understanding of the function and meaning of the objects in their social and cultural context. The method is usable when you want to keep and improve an existing long-lived object such as a building, and it it is very effective in developing a new industrial product on the basis of existing comparable products.
When we invite potential users of a future product to explore and evaluate a product concept or prototype, we often get results quicker if we invite, instead of one person, two friends or acquaintances to "co-discover" and evaluate the product proposal and its potential strong and weak points. Friends usually exchange ideas frankly and openly, therefore their opinions, especially the initial responses to novel products, often are original and innovative. This method has been used by Kemp and Gelderen, p.139, and Jordan 140.
The normal approach when developing a new product is to find an exemplar which can serve as a point of departure for the new object that shall be created. Exemplars are earlier produced artefacts or their details, as evaluated from contemporary point of view and found worth of being reproduced, usually after some improvements to them.
Exemplars can provide useful points of reference in a product design project, particularly in the early phase of preparing a detailed product concept, when it is difficult to find other patterns for describing the future product. A new product idea can be defined either by selecting an exemplar and enumerating the necessary improvements to it, or alternatively by giving a set of exemplars and pointing out the benefits of each, a combination of which should be re-created in the new product. Both these methods are explained in Logical patterns for presenting a product concept. Another phase in a product development project where product proposals are evaluated from a holistic viewpoint is Testing Prototypes. - Also after the fabrication some products are occasionally tested, mostly by consumer organizations and journals, see Public critique of products.
Extensive normative study purports to improve a class of objects. Its result, the normative proposals can be applied to most or all of the objects in this class and perhaps elsewhere, too. In other words, they contain generally valid or "nomothetic" instructions, theory of design for new products. The number of objects in the study will usually be great, and for practical reasons it will be necessary to restrict the amount of information. This is usually done by abandoning the goal of holistic study, and instead selecting for recording and analysis only those attributes of the objects that are important for reaching the normative goal of the project.
In product development, the interesting qualities are mostly related to the requirements that a product is expected to comply with, in other words the goals of product design, such as its usability, function, beauty, meaning, ecology and the environment, economy, or safety.
The methods of descriptive Study of Static Objects can often be used in normative studies, as well, by complementing the measurements with suitable evaluations.
|Tick one box on each line|
Whether this same person then is competent to put on paper the subjective evaluative preferences as well, depends on how the point of view of the project has initially been defined. Most often the selected point of view in extensive study involves so many people that it is not feasible to have all these represented on place. The researcher can then either collect the evaluations from the interest groups separately with suitable interrogating methods, or he can himself do the evaluation on the basis of literature, i.e. earlier research and published theory of design.
The later stages of a normative project are described on the pages Normative Analysis, Evaluating Normative Proposals and Normative Reporting.
August 3, 2007.
Comments to the author:
Original location: http://www2.uiah.fi/projects/metodi