You are on page 1of 20

International Journal of Technology and Design Education (2005) 15:217236 DOI: 10.

1007/s10798-004-1904-4 Springer 2005

Creativity in Design Engineering and the Role of Knowledge: Modelling the Expert
HENRI CHRISTIAANS1 and KEES VENSELAAR2
1

Faculty of Industrial Design Engineering, Delft University of Technology, The Netherlands; E-mail: henri.christiaans@io.tudelft.nl. 2VRC, Utrecht, The Netherlands

ABSTRACT:

The present study focuses on the relationship between the acquisition of design knowledge by novice design students and the quality of their designs. Design learning is typically based on action and reection. Knowledge of solution processes, being part of this reection, is found to be crucial in monitoring and controlling the design process and in reaching an optimal, creative result. The studies described in this article suggest a close relationship between the amount of process knowledge knowledge of managing and monitoring the solution nding process reported by novice designers, and the creativity of the designed product.

Keywords: design learning, design thinking, creativity and cognition, students learning experiences

INTRODUCTION

Design teaching and learning has always received much attention, but contrary to other disciplines research on design education has long been neglected. Although the systematic study of the design process since Eastmans work (1969) has partly been based on students as subjects, the conclusions have mostly focused on design in practice and not on education in practice. Even recent work on the development of design learning strategies and design pedagogy (Eastman et al. 2001) lacks any underpinning by research data. The two main problems the authors encountered were the lack of rigorous evaluations of innovative classroom interventions, and the lack of clarity in determining the gains in student performance on the basis of the learning method applied. Nevertheless, there is consensus in the literature that competence in designing can only be gained through experiencing the design process as a problem-solving event. Practical exposure is essential to the learning of design methods, and the experience should be as real as possible. Novice designers need learning experiences composed of a knowledge component (the representation of the facts, concepts, principles, procedures and/or theories on a certain subject, characterised by learning, remembering and/or reproducing) and a task performance or skills development component (Kirschner et al. 1997). However, academic education in general reects ambivalence

218

HENRI CHRISTIAANS AND KEES VENSELAAR

towards oering students as broad as possible insights into this body of academic knowledge, and (mostly separate) training in practical and professional design skills. While education has an idea about what academic knowledge is relevant to teach, most practical knowledge, however, seems only to be available at an intuitive level, being transferred in an implicit way. What is design education about? It is clear that design education foreshadows the complexity of design problems. First, if learners are to become competent designers they need to gain an understanding of how the various stages of design t together in the design process. They need to progress through the entire event and receive feedback on it. Only after demonstrating competence, can they successfully conduct more complex forms of design. Second, designing asks for the integration of many disciplines such as engineering, aesthetics, psychology and sociology, environmental studies, history, etc. And third, in design education another form of integration is also important; integration between the application of theoretical knowledge and the nal physical embodiment of the design. The experiential learning approach, also characterised in this case as learning-by-doing, is widely accepted as a means of enabling learners to blend theoretical frameworks with real-life experiences (Tynjala 1998). The paradigm underlying this approach is that of discovery learning. It assumes that students in a novel situation will be able to translate that which they have experienced into meaningful insights into the subject matter, and will have both their curiosity aroused and an increased motivation to learn (Pietersen 2002). Project-based learning is one of the more eective methods within the experiential learning approach, as experiential learning activities usually take the form of complex projects that consist of generally structured and guided experiential activities (Tynjala 1998). One could say that experiential learning functions by dual alternation: between action and reection; between unconscious and conscious theories. By engaging both experiential learning functions in a cyclic procedure, students will be able to integrate them. While in most design schools project-based learning has been adopted as the key teaching-learning strategy, the question about the eectiveness of this approach remains unanswered. One of the issues in design education that has only been dealt with at an intuitive level is the enhancement of creative abilities. Creativity is one of the key concepts when judging a designer or a design solution. The main question that nowadays still has to be answered is how can knowledge that enhances creative designing be taught using explicit instruction. Although the importance of intricate and detailed knowledge as being essential to designing is well recognized, little is known about how the knowledge base of the (novice) designer aects the quality or creativity of the design. In order to operationalize the concepts used, we will rst pay attention to what is understood by knowledge base and creativity.

CREATIVITY IN DESIGN ENGINEERING KNOWLEDGE-BASED DESIGN ENGINEERING

219

The term knowledge refers to an individuals personal stock of information, skills, experiences, beliefs and memories (Alexander et al. 1991). When designing multiple knowledge components, both cognitive and aective should interact in order to generate creative products (Alexander et al. 1996). Examples of such conuency theories are the componential model of Amabile (1983, 2001), which distinguishes between domain-relevant skills, creativity-relevant processes and intrinsic task motivation, and the investment theory (Sternberg & Lubart 1996) that includes mood, aect, motivation and personality. In this study we focus on the cognitive components referring to the knowledge base of the designer as the encoded and retrieved design-relevant knowledge, and the strategies that are used to organise and control this knowledge. In search for relevant knowledge in the area of solving ill-structured problems (problems with multiple solutions and several ways of reaching a solution) we have almost only found empirical data that covers the process of solving well-structured problems, such as problems in physics and mathematics (e.g., Larkin et al. 1980; Chi et al. 1981; De long & FergusonHessler 1986; Cooper & Sweller 1987; Sweller 1989; Ericsson & Smith 1991). These and many other studies show that a useful distinction to be made is that between domain-specic knowledge and general process knowledge, coming close to the domain-relevant skills and the creativity-relevant processes of Amabiles componential model. Hence, a generally accepted classication of knowledge is made which distinguishes between domain-specic knowledge and domain-independent general process knowledge. In a study on student learning in architectural design, Venselaar et al. (1987) distinguish two content-related knowledge components within the domain-specic knowledge, together with domain-independent knowledge: Domain-specic basic knowledge in design refers to academic knowledge and skills of dierent disciplines, such as knowledge of mechanics, ergonomics, marketing and psychology, or the skill to operate a sawingmachine. Domain-specic design knowledge refers to knowledge and skills of the design discipline itself; for example, the possibilities of specic materials to attain certain concept solutions, (the application of) design methods, knowledge of existing design solutions, design history, 2D and 3D modelling, aesthetics, and user trials. Together with basic knowledge, design knowledge is specic to a certain domain, in this case industrial design (engineering). General process knowledge refers to domain-independent knowledge of managing and monitoring the solution generating process, part of it being called meta-cognitive knowledge. The overall process involved in solving illstructured problems seems not to be very dierent from the processes found in solving well-structured problems (Simon 1973; Greeno 1980;

220

HENRI CHRISTIAANS AND KEES VENSELAAR

Weisberg 1986; Langley et al. 1987). However, it can be assumed that, because of the need for knowledge from dierent domains and due to the heuristic character of the process, solving ill-structured problems relies heavily on general process knowledge. Knowledge of general methods and heuristics of this kind are denoted weak methods (Anderson 1987; Langley et al. 1987; Rzevsky 1990) as opposed to strong methods, such as the use of algorithms. They help the designer to organise the problem solving process as a whole. It refers to reection on the design process by knowing which stages are relevant in the problem-solving process and which methods can be used to facilitate it. Research on knowledge-based problem solving supports the view that expert problem solving depends primarily on having appropriate domainspecic knowledge and general process knowledge, and not on any unusual intellectual abilities (Anderson 1987; Elio & Scharf 1990). Experienced problem-solvers have an extensive knowledge of problem types and tend to sort problems by solution procedure (Mayer 1987). Together these two dierent knowledge components make up the problem space of the designer, and both are necessary to successfully solve problems.

CREATIVITY AS A CRITERION MEASURE

In most investigations on problem solving, a causal relationship is hypothesized between the problem solving ability of the person, as is determined by his prior knowledge or experience, his strategy during the process, and the quality of the result. Within the domain of design engineering one of the most outstanding criteria for quality is the creativity of the design. But despite the enormous amount of research on creativity in the last decades it is very dicult to select an operational denition that is relevant for design education and, more importantly, one that leads to objective methods for assessment. In the literature, denitions are given in terms of a personality trait, the process itself or the product (Hedge & Lawson 1979). For educational purposes a focus on the trait approach is not very interesting, since a trait by denition is a very stable disposition and therefore not easily inuenced by training. The process-approach gives a valuable starting point for analyzing the process itself, but fails also in oering a valuable criterion. Therefore the outcome of the process, the product or artifact, seems to be the only result that is accessible for assessment. Amabile (1983) has developed an elegant strategy for the assessment of creativity. In her opinion creativity can be seen as a property of products. A product is creative to the extent that observers, familiar with the domain of activity, independently agree it is creative. Her studies show that if such observers independently judge artistic products, the inter-judge-reliability is high. The results of this study were conrmed by a study of Christiaans (1992, 2002a). If it is true that creativity of the products of design engineering can be assessed reliably by this so-called consensual assessment technique, it is then interesting to search for the relationship

CREATIVITY IN DESIGN ENGINEERING

221

between creativity of the product and the knowledge applied by the designer to attain that result.

AIMS

In order to improve education within design engineering, attention should be paid both to domain-specic knowledge of relevant domains, and domain-independent general process knowledge necessary in process management and control. The student has to learn to reect on what he is doing whilst solving a design problem, and to understand the process and strate gies underlying the most ecient way of reaching that solution. Since Gagne (1970) states that the outcome of the learning process is the acquisition of higher-order rules that are capable of being generalized in a wide variety of stimulus situations, an increase of this general process knowledge during the learning process may be expected. Implicit in this statement is the claim that an increase in mastering higher-order rules improves problem solving. Will this hold for design learning as well? The objectives of this study are to determine: 1. What knowledge is encoded, retrieved and applied by a student during the design process? 2. What knowledge is applied to manage the problem-solving process? 3. To what extent does this knowledge management aect the resultant design creativity?

STUDY I

In order to investigate the aforementioned aims, the purpose of study I is to explore: (1) the kind of knowledge students acquire in the rst-year design course, and (2) the relationship between knowledge acquisition by novice design students throughout the year and their design creativity. The design course can be characterized as project-based learning, an eective method within the experiential learning approach. This course is part of the rst-year program, runs throughout the year and accounts for 170 h of a total of 1700 h. The course encompasses four design projects of dierent lengths (see Table I). Project I assignment: design a waste basket for a specic situation. Project II assignment: design a tool for heating food in the open air (a form of barbecue). Project III assignment: design a public telephone booth or a computer cabinet. Project IV assignment: design play-equipment for 48 year-old children. Students work either individually (project III) or in groups of 46 students (all other projects), the common group size for design projects in the rstyear (the composition of groups is dierent for each project).

222

HENRI CHRISTIAANS AND KEES VENSELAAR


TABLE I Some characteristics of rst year design projects

Project

Semester 1 First Quarter I II 4 1 g end 16 Second Quarter III 6 2 i end 48

Semester 2 Third Quarter IV 6 2 g halfway 48 6 2 g end 48 Fourth Quarter

Weeks Sessions/week Group/Individual Learner report Hours in total

2 1 g end 8

Method Subjects Subjects were 20 rst-year students from the School of Industrial Design Engineering (IDE) at Delft University, the Netherlands, they were randomly selected from the total rst-year population. Learner reports Students retrospective reports were used as data input for this study. During the design course, students were asked to write a learner report at the end of a design project. Due to the length of project IV, learner reports were also written halfway through the project. This method, introduced by De Groot (1974), implies that the problem solver writes his or her learning experiences over the preceding period, starting each experience with: I have learned (that or how) De Groot developed a theory on the assessment of educational learning objectives using the learner reports as data. In our study the learner reports throughout the year, provide logitudinal data on knowledge acquisition. Two judges (the authors of this article) independently rated all learner reports by assigning each learning experience to one of the categories basic knowledge, design knowledge and general process knowledge. Each of the judges rated over 700 learning experiences. The agreement between the judges will give relevant information about the reliability of the assessment. The measure used for this inter-rater agreement is Cohens kappa (Cohen 1960). Design quality assessment As three of the four projects are team based, the results could not be considered in order to measure a students individual result. Only the products designed individually in project III were used. These products were judged on creativity according to Amabiles consensual assessment technique.

CREATIVITY IN DESIGN ENGINEERING

223

Each judge individually rates each design on a number of attributes such as creativity and technical quality. The judges were 10 graduate IDE students. In a previous study it was shown that in judging products on creativity, the inter-rater reliability among selected graduate design students was as high as that of design teachers (Christiaans 2002b). The designs were modelled in coloured or plain cardboard and were 15 20 inches high. The models were photographed individually in two dierent positions and recorded on slides. These slides were used in this study. Results The pictures in Figure 1 show examples of the design results for a telephone booth. Learner reports An impression of some learning experiences are presented in Table II. The analysis of learner experiences by the two judges gives the following results (see also Table III): The proportion of ratings on which the two judges agree is .64. Coecient Kappa (j) (Cohen 1960) for inter-rater agreement is j=.51. As this agreement is not very high, it was decided only to use the learner experiences that the judges agreed upon for further analysis (the values shown on the diagonal of Table III). Knowledge development through the year The development of students knowledge acquisition during their rst year can be monitored through learning experiences. Using the knowledge components (basic, design and process knowledge) this development is presented in Figure 2. The data shows that domain-specic basic knowledge is mentioned most frequently early in the year and decreases towards the end, while both design knowledge and general process knowledge show a signicant growth during the year.

Figure 1. Models of a telephone booth, the 3rd design project in the rst year.

224

HENRI CHRISTIAANS AND KEES VENSELAAR

TABLE II Examples of learning experiences for each category (1:domain-specic basic knowledge; 2:domain-specic design knowledge, 3:general process knowledge) I have learnt about wood linking and construction. how to handle machines and manufacturing methods. if the paint is going to ake, then use paint which adheres better. that I have to keep up my report by making notes throughout the process. Otherwise I forget the details. that in this design task much attention must be paid to constraints and specications. how to pose a design problem, to make a list of requirements, to describe the users and to write a chronological design report. that the use of colours can clarify chaotic situations and that colours can give meaning to your design. before making the scale model I have to check if the required materials are available (in stock). that the design process is made up of an alternation between creative and exact stages. that if the function of the object is ambiguously stated, then you have to dene the problem clearly yourself and emphasise specic operation by drawing the concepts. if you get stuck within the idea-generation, try to start from totally dierent points of view in order to create new ideas. that gathering information and determining the requirements before generating ideas is very ecient. Cat. 1 1 1 1 2 2 2 2 3 3

TABLE III Ratings for knowledge categorisation of learner experiences by two judges. n = 20 (students), k = 730 (learning experiences). The experiences the raters agreed upon are on the diagonal Rater 2 Knowledge Rater 1 Basic Design Process TOT Basic 229 55 29 313 Design 42 95 71 208 Process 38 21 150 209 Tot 309 171 250 730

Design quality Creativity measurement was based on the artifacts designed in design project III of the rst year. The artifacts are telephone booths and computer desks. The inter-rater reliability (coecient alpha) amongst the senior design student-judges for the two designs was .81 and .86, respectively. In order to

CREATIVITY IN DESIGN ENGINEERING

225

Figure 2. Knowledge acquisition throughout the year.

give each design a score for creativity a Product Creativity Rating (PCR) was created by averaging the creativity judgements per artifact and thus per student. Relationship knowledge components and design creativity A score was determined for each subject for the three knowledge components (basic, design and process knowledge). These scores were the sums of the ratings of all projects through the year. Only the ratings agreed by the two judges were taken into account. The product-moment correlations between each knowledge component and the PCR are presented in Table IV. In order to answer the question whether a possible relationship is really dependent on creativity, the

TABLE IV Product-moment correlations between knowledge components and product creativity (n = 20; nHL = 14). Knowledge components + Basic PCR knowledge Basic knowledge Design knowledge Process knowledge PCR (crea) PCRHL
*

Design knowledge

Process knowledge

PCR

HL PCR

.41 ).26 ).15 ).29 ).01 ).23 ).29 .44* .50*

.90**

p < .05** p < .01

226

HENRI CHRISTIAANS AND KEES VENSELAAR

correlation was also measured between knowledge components and the extreme PCR-scores in the sample (HighLow, HL). Average PCRscores 6.5 and < 5 were used as HL-criteria. Based on this selection, only 14 subjects could be given a HL-score. The ndings show a signicant correlation between creativity and process knowledge Two ndings are of special interest. First, the correlation of .44 and .50 between creativity and process knowledge indicates that, according to the reports, subjects who designed the most creative products acquired more process knowledge than other students. This nding could indicate that designing more creative products is partly dependent on the application of process knowledge. Second, the correlation of .41 (p .07) between basic knowledge and design knowledge is striking. This nding, together with the negative correlation between both domain-specic knowledge components and general process knowledge, supports the hypothesised distinction between the two. Conclusions Learner reports are shown to be an eective learning tool for students, helping them to reect on their own learning process (Van Eyk 1982). In this study the learner reports were used as longitudinal data for the analysis of knowledge acquisition and the relation with design creativity. The results show the dynamics of knowledge acquisition during the year. Regarding domain-specic knowledge, the reported learning of basic knowledge decreases throughout the year, while design knowledge and domain-independent process knowledge increase. The data seem to conrm the assumption expressed previously, that, by performing design projects during the year, students acquire general process knowledge, heuristics and higher-order rules that can be transferred to a wide variety of design situations. Hence, in their learning experiences students more and more stress the acquisition of this kind of knowledge, while the amount of basic knowledge in their learner reports decreases. This rst study was meant to explore knowledge categories applied in solving design problems. A relation has been found between general process knowledge and creativity. Creativity in this study was measured using the years individually designed product, i.e. project III (the designs of the other projects were the result of team work). Therefore the question to be asked is: What will the relation be between the data on knowledge and creativity in every separate design project? Study II was designed to answer this question.

STUDY II

The purpose of study II is to nd evidence for the explorative data of study I. Again, the aims of this second study are to determine: (1) The use

CREATIVITY IN DESIGN ENGINEERING

227

of learner reports in determining the knowledge acquisition of novice design students, and (2) the relationship between knowledge acquisition and creativity ratings. In this study data were again gathered from rstyear design students, however in this case from a more recent population. Due to curriculum reorganisation, three out of the four design projects were performed individually, enabling the relationship between creativity and knowledge acquisition to be studied separately for these three projects. Method Subjects Subjects were rst-year IDE students from Delft University. A group of 55 students was randomly selected out of the total rst-year population of 240 students. A major problem with experiments conducted within a real university course is the signicant drop in student numbers during the year. This also happened in this study. In design project III n 55, in project IV n 35, and in project V n 30. Design course The assignments for the design projects in this year were: a bicycle luggage carrier (project I), a Barbecue (project II), a Telephone Booth (project III), a Shop-window Display (project IV), and a Drill Holder (project V). The products resulting from Project I (taking only half a day), and project II (being a team project), were not included in the creativity assessment. However, the learner experiences of the whole year, including those written at the end of projects I and II were taken into account. Learner reports Students in the sample were asked to write their learning experiences midway (except for project I) and at the end of each project. The midway report was written after the conceptual stage, including problem nding and idea generation. The end-report was concerned with design elaboration: making technical and presentation drawings, constructing a prototype and nishing the design report. The same instructions for writing a learner report were used as in study I. The two judges (the authors) rated all learner reports using the categorisation according to the knowledge matrix. However the rating method diered substantially. The assistance of a computer program was enlisted in order to increase the objectivity of the procedure. The computer program randomly selected one experience at a time, taken from any subject, and displayed it on the screen. After the judge had coded this experience the next one was presented and so on. In this way the judges did not know from which learner report the experience was selected.

228

HENRI CHRISTIAANS AND KEES VENSELAAR

Design creativity assessment In assessing the quality of student work, only the outcomes of the individual projects were taken into account. Creativity was again the criterion measure. Scores for this attribute were obtained from design projects II IV Judges were 10 senior design students who independently gave ratings to the objects. As in study I, the designs were presented on slides. The designs of a telephone booth and a shop-window display were modelled in coloured or plain cardboard about 1520 inches high. The drill holder project did not result in a model so two slides for each design were presented: a rendering (presentation drawing) of the design and a technical drawing. Results Figure 3 shows examples of the design results for a drill holder (project V). Learner reports The analysis of the learning experiences by the two judges gives the following results: The proportion of ratings in which the two judges agree is .74. Coecient Kappa(j) (Cohen 1960) for inter-rater agreement is j .60. This is higher than in study I. However, as in study I, only learner experiences agreed by the judges were used in the analysis. The ratings are reasonably well distributed over the three knowledge components (see Table V for the ratings of one of the design projects). Knowledge development during the year Figure 4 shows the development of three knowledge components during the year. Similar to study I basic knowledge is mentioned most frequently early in the year and decreases toward the end. Design knowledge increases gradually, while general process knowledge shows an initial increase with a drop in the last project.

Figure 3. Renderings of a drill holder, the 5th design project in the rst year.

CREATIVITY IN DESIGN ENGINEERING

229

Table V Ratings for knowledge components of learner experiences by two judges; project III, midway. k = 472 (learning experience). The values on the diagonal (bold) are the experiences the raters agree upon. Rater 2 Rater 1 Knowledge Basic Design Process Total Basic 178 24 7 209 Design 17 48 10 75 Process 29 35 124 188 Total 224 107 141 472

Design creativity assessment The reliability coecients (alphas) for the creativity ratings of the 10 judges are higher than in the rst study, ranging from .89 to .93. In order to give a score for the creativity of each subjects designed product, a PCR was dened by averaging the creativity scores of the 10 judges. Relationship knowledge components and creativity Product-moment correlations between the three knowledge components and the PCR were measured separately for projects IIIV. As in study, the correlation was also measured between knowledge codes and the extreme PCR-scores (HighLow). The average PCR-scores 6 and 4.5 are used as HL-criteria. See Table VI. There is a positive correlation between creativity and domain-independent process knowledge through all projects. In two of the projects this

Figure 4. Knowledge acquisition of novice design student through the year.

230

HENRI CHRISTIAANS AND KEES VENSELAAR

correlation is signicant. Secondly, contrary to the result of study I, there is a negative correlation between basic knowledge and design knowledge in two of the projects. Examples of learning experiences regarded as general process knowledge by students with a high PCR are included in Table VII. Conclusions Results of this second study on knowledge analysis partly supports the results from study I. Once again, a strong positive correlation was found between the product creativity of students and their general process knowledge gained from three projects reported throughout the year. Comparison of the acquisition of general process knowledge shows similar lines in the two samples. During the year there is a gradual increase, although in the sample of study II, the last design project shows a dramatic drop in general process knowledge. One explanation could be that in this last project there is more emphasis on the materialising element of the design process. The negative correlation (in two of the three projects) between basic

Table VI Product-moment correlations between knowledge components and Product creativity Knowledge components + PCR Basic Design Process PCR knowledge knowledge knowledge HLPCR

(a) Project III, telephone booth (n = 55, nHL = 40) Basic knowledge Design knowledge .10 Process knowledge ).21 ).04 PCR (crea) .23 ).20 PCRHL ).10 ).25 (b) Project IV, window display (n = 35, nHL = 18). Basic knowledge Design knowledge ).43* .10 Process knowledge ).33* ).03 PCR (crea) ).40* HL PCR ).24 ).52* (c) Project V, drill holder (n = 30, nHL = 23). Basic knowledge Design knowledge .29 Process knowledge ).19 .13 PCR (crea) ).06 .05 HL PCR .08 .13
*

.14 .20

.94**

.31 .45*

.93**

.39* .47**

.90**

p < 0.05,

**

p < 0.01.

CREATIVITY IN DESIGN ENGINEERING

231

knowledge and design knowledge is puzzling. Whilst in the rst study a logical explanation for the positive correlation was found in the assumption that both were domain-dependent knowledge components, this explanation has now been invalidated. On the one hand the correlations in this study justify the distinction made between basic knowledge and design knowledge. Designing asks for the application of knowledge from other disciplines such as knowledge of materials, ergonomics, mechanics, etc. Design knowledge

TABLE VII Examples of learning experience regarded as general process knowledge In order to be successful I should always be open-minded to changes/improvements in my design; the outer appearance of the design will also change all the time. I have learned that designing is not only restricted to the studio hours, but that it keeps me busy 24 h a day. Look around you and you learn as in a ow production. All artifacts have already been designed once. And now that I understand a bit what designing is about, I experience designs more consciously. Watch how they have solved certain problems. Ive got a better understanding of the whole design process. It was the rst time I had to make a serious concept choice, a detailed elaboration of the concept selected, a logical report and a presentation report. It was really good that we were forced to take the work serious while nothing happens when you dont do it. In that way I could gure out how to do it and, mostly, how to do it next time. This nal design meets the requirements stated before. Now its time to observe the environment and to synthesize these observations to principles which might be applied to future design assignments. I have learned that its possible to develop a lot of assessment systems which will lead to dierent choices. Most important in my opinion is that beforehand I have to develop a well-considered system. With the system I will explain why I select in this way and how the assessment system works. After I will do the selection in order to attain a rational choice. I have seen that I shouldnt put up myself with a solution which is OK but not optimal. When I consider it for a while (could be a couple of hours) the previously impossible appears to be possible. I have learned a lot about myself. For example that at the beginning of the project I am not concentrated yet on ideas which are then generated; although they at that stage are not so realistic, later on they often turn out to be very useful. And when half-way the project I am at full swing, I have to work day and night; but thats what I like. In this project I have learned that making and criticizing a model gives an understanding of the design. Thought mistakes and construction errors become obvious very quickly. Its easy to evaluate the design on the basis of the model.

232

HENRI CHRISTIAANS AND KEES VENSELAAR

itself is the ability to integrate knowledge from these discipline and at the same time knowledge of precedents, of how to conceptualise ideas and to give a physical form to those ideas. On the other hand it is dicult to explain why the correlation in study I diers so much from the ones in study II.

GENERAL DISCUSSION

The purpose of design education is to transfer knowledge of solving design problems in such a way as to prepare students for their development from novices to design experts. As a matter of fact, little is known about the way in which knowledge and skills within this domain are actually acquired and used. The aim of this study was to nd explicit, detailed knowledge of the design engineering process as a whole, paying special attention to productand process-related knowledge. An understanding of the kind of knowledge a designer needs during the design process would be an important contribution to problem solving in general and design education in particular. The next step would be to develop the correct learning situation in order to transfer this knowledge eectively and eciently. An understanding of the design engineering process was operationalised in this study in terms of knowledge components as the input for the design process, and of creativity as an expression of the output, the design quality. Research on this topic within the domain of industrial design engineering is sparse. Creativity was measured with high values of interrater reliability, hence being a useful criterion in measuring the perceived quality of products. In the next step attention was paid to the measurement of knowledge used during the design process. An elicitation method was introduced based on the retrospectively written learner reports of rst-year students. These reports were categorized in one of three knowledge components: domainspecic basic and domain specic design knowledge, and domain-independent process knowledge, the latter being a more general, meta-cognitive, knowledge component of value in ill-dened problem solving processes. Although the interrater reliability was not high, there was a substantial amount of learning experiences (about 50%) on which there was agreement. Based on coding by the two judges on these agreed learning experiences reliable conclusions could be drawn with respect to the development of knowledge acquisition during the year, and to the relationship between this knowledge acquisition and the creativity of the design solutions. Two studies were conducted. In study I knowledge elicited through the year was correlated with the product creativity rating of just one project, while in study II correlations were measured between knowledge elicited in one project and product creativity ratings in the same project. First, the data show that during the year, students increasingly emphasise the learning of design knowledge and general process knowledge at the expense of domain-specic basic knowledge. This result is similar in both

CREATIVITY IN DESIGN ENGINEERING

233

studies. It is fair to conclude that by working on design projects during the year, and by repeating the same kind of procedures, students build up their knowledge of designing and the supporting processes in general, in terms of heuristics and higher-order rules. Second, in almost all separate design projects in both studies, a relationship was found between the creativity of the design and the amount of process knowledge elicited. There is a consequent tendency for subjects whose designs have a higher creativity rating to elicit on average a greater amount of general process knowledge than other students. What is the meaning of this relationship? Process knowledge was dened as domainindependent knowledge related to the monitoring function of memory that helps the problem solver to organize and evaluate the problem-solving process as a whole. A tentative conclusion could then be that students who have acquired this process knowledge, and are guided by it (as suggested by the fact that they are able to reect on this kind of knowledge in their learner reports), are better able to generate creative solutions. Knowledge of weak methods is possibly of importance during the rst-year as students lack an adequate knowledge base; strong methods, if applicable at all in the domain of designing, do not yet belong to the repertoire of novice designers. The tendency could also mean that novice designers who design creative products have already acquired certain expert qualities; i.e. they possess wellorganized abstract knowledge in terms of underlying principles (Anzai 1991). However, this explanation has been contradicted by the learning curves in the two studies, which both start with a low level of process knowledge. Another hypothesis is that better designers are better able to reect consciously on their acquired domain-independent knowledge. A counter-argument here being, that in these studies the same students do not design creative products all the time; for example the product of a student designed in project 1 is rated as very creative, while the products of the same student in projects 2 and 3 get an average score. Hence, the relationship between creativity and amount of general process knowledge is not only person-related (Simonton 2000). This corresponds with ndings in other studies (see Sternberg & Lubart 1996) that there are moderate correlations between product creativity ratings within the same domain. The conuence approach presents an interesting point of view, which presumes that several factors such as motivation, perseverance, intellectual risk taking, knowledge of the domain and the like must interact in such a way that an optimal design is reached by the designer (Sternberg & Lubart 1996). Mastering general process knowledge could be one of those factors. One of the questions to be answered by this study is: What kind of curriculum the novice designer should follow in order to become an expert, or in terms of the present study: how to learn process-knowledge? One of the main problems is how this knowledge can be explicitly addressed. In design research this topic, which also refers to the notion of tacit knowledge, is recently under discussion again also inuenced by the renewed interest in Schons work (1983). Halpern (1998) proposes questions that should be

234

HENRI CHRISTIAANS AND KEES VENSELAAR

asked as a way of meta-cognitive monitoring, in order to enhance critical thinking in problem solving, which seems to be very close to creative thinking. Examples of such questions are: state the problems in at least two ways, list two solutions to the problem, and: present two reasons that support the conclusion and two reasons that do not support the conclusion. In fact, students have to learn to state goals, have to know their departure point, and have to realize throughout the design process exactly where they are in the process of reaching that goal. These processes should be made more explicit in order to generate possibilities for students to transform themselves from novices into experts. The overall conclusion is that the method of knowledge measurement deserves further development. First, a better denition of knowledge categories, together with a simpler way of encoding would improve the current classication. Second, the method of eliciting knowledge from (novice) designers using retrospective reports leaves much to be desired. Another elicitation technique for further research should be recommended, so that there can be a more direct denition of what the designer is doing, and what knowledge is being used. Finally, the studies described here provide some interesting, but speculative, results with regard to the development of knowledge acquisition during the year, and to the relationship between knowledge and performance quality. As Anderson states in his theory of complex cognition, the power of human cognition depends on the amount of knowledge encoded and the eective deployment of the encoded knowledge (Anderson 1996, italics added by present authors). The study presented here highlights the importance of knowledge of eective deployment methods.

REFERENCES
Akin, O.: 1986, Psychology of Architectural Design, Pion Limited, London. Alexander, P. A. & Judy, J. E.: 1988, The Interaction of Domain-Specic and Strategic Knowledge in Academic Performance, Review of Educational Research 58, 375404. Alexander, P. A., Schallert, D. L. & Hare, V. C.: 1991, Coming to Terms: How Researchers in Learning and Literacy Talk about Knowledge, Review of Educational Research 61, 315343. Alexander, P. A., Jetton, T. L. & Kulikowich, J. M.: 1996, Interrelationship of Knowledge, Interest and Recall: Assessing a Model of Domain Learning, Journal of Educational Psychology 87, 559575. Amabile, T. M.: 1983, The Social Psychology of Creativity, Springer-Verlag, New York. Amabile, T. A.: 2001, Beyond Talent: John Irving and the Passionate Craft of Creativity, American Psychologist 56(4), 333336. Archer, B.: 1963, Systematic Method for Designers, Design 4. Anderson, J. R.: 1987, Skill Acquisition: Compilation of Weak-method Problem Solutions, Psychological Review 94, 192210. Anderson, J. R.: 1996, ACT: A Simple Theory of Complex Cognition, American Psychologist 51 (4), 355365. Anzai, Y.: 1991, Learning and Use of Representations for Physics Expertise, in K. A. Ericsson & J. Smith (eds.), Toward a General Theory of Expertise pp. 6492, Cambridge University Press, New York.

CREATIVITY IN DESIGN ENGINEERING

235

Bransford, J., Sherwood, R., Vye, N. & Rieser, J.: 1986, Teaching Thinking and Problem Solving, American psychologist 41, 10781089. Chi, M. T., Feltovitch, P. J & Glaser, R.: 1981, Categorization and Representation of Physics Problems by Experts and Novices, Cognitive Science 5, 121152. Christiaans, H. H. C. M.: 1992, Creativity in Design. The Role of Domain Knowledge in Designing, Lemma B.V., Utrecht. Christiaans, H. H. C. M.: 2002b, Creativity as a Design Criterion, Creativity Research Journal 14, 4154. Christiaans, H. H. C. M.: 2002a, Book Review of C. Eastman, M. McCracken, & W. Newstetter (eds.), (2001). Design Knowing and Learning: Cognition in Design Education. Elsevier, Oxford. Design Studies 23(4) 2002, 433434. Cohen, J.: 1960, A Coecient of Agreement for Nominal Scales, Educational and Psychological Measurement 20, 3746. Cooper, G. & Sweller, J.: 1987, The Eects of Schema Acquisition and Rule Automation on Mathematical Problem-solving Transfer, Journal of Educational Psychology 79, 347362. Eastman, C. M.; 1969, On the Analysis of Intuitive Design Processes, in G. T. Moore (ed.), Emerging Methods in Environmental Design and Planning, The MIT Press, Cambridge, Massachusetts. Eastman, C., McCracken, M. & Newstetter W.: (eds.) (2001). Design Knowing and Learning: Cognition in Design Education, Elsevier, Oxford. Elio, R. & Scharf, P. B.: 1990, Modelling Novice-to-expert Shifts in Problem Solving Strategy and Knowledge Organization, Cognitive Science 14, 579639. Ericsson, K. A. & Smith, J.: 1991, Toward a General Theory of Expertise, Cambridge University Press, New York. Eyk, G. H. A. van.: 1982, The Writing of Learning Experiences as a Teaching Tool, in R. Langdon, K. Baynes & P. Roberts (eds.), Design Education: Vol.5. Proceedings of the Design Policy Conference, London, 1982, Design Council, London. Gagne, R. M.: 1970, The Conditions of Learning (2nd ed.), Holt, Rinehart & Winston, London. Greeno, J. G.: 1980, Trends in the Theory of Knowledge for Problem Solving, in D. T. Tuma & F. Reif (eds.), Problem Solving and Education pp. 925, Erlbaum, Illsdale, NJ. Groot, A. D. de: 1974, To What Purpose, to What Eect? Some Problems of Method and Theory in the Evaluation of Higher Education, in W. A. Verreck (ed.), Methodological Problems in Research and Development, Swets & Zeitlinger, Amsterdam. Halpern, D. F.: 1998, Teaching Critical Thinking for Transfer Across Domains: Dispositions, Skills, Structure Training and Metacognitive Monitoring, American Psychologist 53(4), 449455. Hedge, A. & Lawson B. R.: 1979, Creative Thinking, in W. T. Singleton (ed.), Compliance and Excellence. The Study of Real Skills, pp. 280305, MTP press ltd. Lancaster. Jong, T. de & Ferguson-Hessler, M. G. M.; 1986, Cognitive Structures of Good and Poor Novice Problem Solvers in Physics, Journal of Educational Psychology 78, 279288. Kirschner, P., Van Vilstercn, P., Hummel, H. & Wigman, M.: 1997, The Design of a Study Environment for Acquiring Academic and Professional Competence, Studies in Higher Education 22, 151171. Langley, P., Simon, H. A., Bradshaw, G. L. & Zytkow, J. M.: 1987, Scientic Discovery. Computational Explorations of the Creative Processes, The MIT Press, Cambridge, MA. Larkin, J., McDermott, J., Simon, D. & Simon, H.: 1980, Models of Competence in Solving Physics Problems, Cognitive Science 4, 317348. Mayer, R. E.: 1987, Educational Psychology: A Cognitive Approach, Little, Brown, Boston. Pietersen, C.: 2002, Research as a Learning Experience: A Phenomenological Explication, The Qualitative Report 7(2). http://www.nova.edu/ssss/QR/QR7-2/pietersen.html. Schon, D. A.; 1983, The Reective Practitioner, Basic Books, New York. Simon, H. A.: 1973, Structure of Ill Structured Problems, Articial Intelligence 4, 181201. Simonton, D. K.: 2000, Creativity: Cognitive, Personal, Developmental and Social Aspects, American Psychologist 55(1), 151158. Sternberg, R. J. & Lubart, T. I.: 1996, Investing in Creativity, American psychologist 677688.

236

HENRI CHRISTIAANS AND KEES VENSELAAR

Sweller, J.: 1989, Cognitive Technology: Some Procedures for Facilitating Learning and Problem Solving in Mathematics and Science, Journal of Educational Psychology 81, 457 466. Tynjala, P.: 1998, Traditional Studying for Examination Versus Constructivist Learning Tasks: Do Learning Outcome Dier?, Studies in Higher Education 23, 173189. Venselaar, K., Gerritsen van de Hoop, W. & Drunen, P. van: 1987, The Knowledge Base of the Designer, in P. R. J. Simons & G. Beukhof (eds.), Regulation of learning, pp. 121135, SVO, The Hague. Weisberg, R W.: 1986, Creativity. Genius and Other Myths, Freeman, New York.

You might also like