Elsevier

Design Studies

Volume 30, Issue 1, January 2009, Pages 38-59
Design Studies

Exploring problem decomposition in conceptual design among novice designers

https://doi.org/10.1016/j.destud.2008.07.003Get rights and content

Conceptual product design is commonly described as problem solving. In the present study we attempt to expand this view. Focusing on the solution search phase, we analyse explicit and implicit problem decomposition techniques and integrate them into a descriptive cognitive model. To evaluate the prevalence of decomposition modes empirically, we provide results from a verbal protocol analysis study involving 16 senior students of mechanical engineering. Data indicated that the subjects apply top–down control strategies coupled to implicit decomposition. Explicit decomposition was used seldom and without obvious benefits. We relate these results to the model that considers implicit decomposition as an integral part of the problem interpretation process and discuss the role of decomposition in a structured idea generation process.

Section snippets

Design cognition

The current investigation approaches decomposition from the perspective of design cognition (Eastman, 2001). This paradigm describes design as information processing, employing explanatory concepts such as problem spaces, goal hierarchies, control strategies, and search methods (Akin, 1986, Chandrasekaran, 1990). In this framework, design is considered as ill-structured problem solving (Simon, 1973) because design problems define the problem in a fuzzy manner, with regards to initial state,

Empirical study

After outlining a model of design IG, we next present a re-analysis of a large dataset (originally reported in Perttula and Liikkanen, 2006) to examine the relationship of implicit and explicit decomposition in conceptual product design. The main interest is in how novice designers use explicit and implicit decomposition in the design IG process and whether the process can be described by our IG model. To this end we employ verbal protocol analysis, which is a widely used method among design

Results

All subjects were able to produce several alternative concepts during the experiment (M = 9.8, SD = 3.6 concepts). More ideas were generated in the Forest than in the Plant task (10.6 vs. 8.9 concepts, respectively), but the difference was not statistically significant (T(7) = 1.10, p > 0.10). Subjects verbalized a considerable amount of information while working (M = 869.9, SD = 378.6 words per protocol) and the data from all designers were included in the analysis.

Designers talked mostly about their

Discussion

In this paper we have developed the theory of problem decomposition in conceptual product design and empirically reproduced some previous findings (Ho, 2001). We proposed implicit decomposition as an instinctive technique for interpreting familiar problems. The results showed that the working model of design IG was useful in exploring decomposition through the design protocols. For instance, the processes of recognition and analogical inference were utilized in the task that concerned a

Conclusions

In this paper, we attached a dual-stage view of problem decomposition into a model of design IG and used the model to describe verbal protocol data from 16 designers. It appeared that explicit decomposition was an unnecessary step in the IG process as designers successfully implicitly decomposed the problem. This fits our model which proposes implicit decomposition as a part of the problem interpretation process. This result also reflects the nature of design as an implicitly controlled

Acknowledgements

This work was conducted as a part of a research project funded by the Academy of Finland. We thank Miikka Vanhamaa for providing the task brief illustrations, J. Matias Kivikangas for transcribing the protocols, and all the people whose comments helped us to improve the manuscript, especially the two anonymous reviewers.

References (53)

  • P. Lloyd et al.

    Discovering the design problem

    Design Studies

    (1994)
  • H.A. Simon

    The structure of Ill structured problems

    Artificial Intelligence

    (1973)
  • Ö Akin

    Psychology of architectural design

    (1986)
  • C. Alexander et al.

    A pattern language

    (1977)
  • J.R. Anderson

    The architecture of cognition

    (1983)
  • L.J. Ball et al.

    Cognitive processes in engineering design: a longitudinal study

    Ergonomics

    (1994)
  • L.J. Ball et al.

    Problem-solving strategies and expertise in engineering design

    Thinking and Reasoning

    (1997)
  • L.F. Barrett et al.

    Individual differences in working memory capacity and dual-process theories of the mind

    Psychological Bulletin

    (2004)
  • B. Chandrasekaran

    Design problem solving: a task analysis

    AI Magazine

    (1990)
  • H. Coskun et al.

    Cognitive stimulation and problem presentation in idea-generating groups

    Group Dynamics: Theory Research and Practice

    (2000)
  • N. Cross et al.

    Analysing design activity

    (1996)
  • A.R. Dennis et al.

    Structuring time and task in electronic brainstorming

    MIS Quarterly

    (1999)
  • A.R. Dennis et al.

    Process structuring in electronic brainstorming

    Information Systems Research

    (1996)
  • K. Duncker

    On problem-solving

    (1945)
  • C. Eastman

    New directions in design cognition: studies of representation and recall

  • D.E. Egan et al.

    Theories of rule induction: knowledge acquired in concept learning, serial pattern learning, and problem solving

  • Cited by (78)

    • FloorplanGAN: Vector residential floorplan adversarial generation

      2022, Automation in Construction
      Citation Excerpt :

      This suggests that FloorplanGAN can be further developed into interpretable models and has a potential to be integrated with human-machine collaborative design process. Meanwhile, residential floorplan generation is not a task that can be easily compared quantitatively with other methods because of the open-ended and ill-defined nature of architectural design problem [39]. The evaluation of architectural floorplans is a sophisticated issue, with both subjective and objective criteria.

    • Command prediction based on early 3D modeling design logs by deep neural networks

      2022, Automation in Construction
      Citation Excerpt :

      Therefore, methods for data preparation and data mining based on BIM logs especially on early design stages has become a key area to study [17]. In the early design stage, there are many design changes and design adjustments due to the open-ended and ill-defined nature of design problems [48]. Since architecture is made up of different building components, designers constantly make changes and adjustments to same functional components during the design phase.

    • A data structure for studying 3D modeling design behavior based on event logs

      2021, Automation in Construction
      Citation Excerpt :

      This case study focused on problem-solving behavior in 3D modeling design. In design cognition, decomposition of design problem into sub-problem was widely observed and was expressed as “a weak method” for designers to deal with the ill-defined problems [16–18]. Designers are more solution-oriented comparing to other professionals [19].

    • A novel user-based gesture vocabulary for conceptual design

      2021, International Journal of Human Computer Studies
    • Towards an integrative approach to researching design expertise

      2021, Design Studies
      Citation Excerpt :

      Chen and You (2004) found that novices' lack of knowledge impeded their development of merely sound solutions, while experts produced more detailed and complex solutions than the average practitioner. However, as studies by Adams et al. (2011) and Liikkanen and Perttula (2009) demonstrated, novice designers' lack of knowledge does not necessarily prevent them from solving ill-defined problems. The results of these studies stand in contrast to the skill acquisition model, because they show us that novices often develop much more creative and unusual solutions than experts, even if these outcomes are not necessarily as practical in terms of the cultural expectations of clients.

    View all citing articles on Scopus
    View full text