Abstract

There are two main purposes of this thesis. First, it refines existed EBD coding scheme which used to be called formal model of EBD process and develops a three-layer structure of EBD coding scheme. The refined EBD coding scheme is expressed as hierarchical structure. It is composed by three levels. The first level includes collecting, processing and expressing. The second level describes the design process as five basic parts: identifying problems, searching information, generating solutions, expressing solutions and evaluating solutions. The third level is formed by 7 parameters: analyzing problems, identifying conflicts or (new) requirements, searching synthesis knowledge, searching evaluation knowledge, generating solutions, expressing solutions and evaluating solutions.
Second, it focuses on applying this coding scheme in a case study to find out the relationship between the level of details and uncertainties in design research. Through hypothesis test, the conclusion shows that the uncertainty increases when the level of details is deeper under the fixed technique. In this thesis, a designer was assigned to solve a design task, and the whole design process was recorded as a verbal protocol. The subjects apply the EBD coding scheme to define the actions of the design.
This coding scheme has several benefits. First, this coding scheme is a generic and logical. In the study, five subjects were chosen to do the experiments. Via EBD coding scheme, the participants coded all the segmentations of the design process step by step. Second, the refined scheme is apt to understand and convenient for manipulating. The definition of the parameters of the EBD coding scheme is based on the previous researches and sources. No matter the subjects from which kind of background, through short-term trainings and continuous practice, they can operate the scheme easily and smoothly. The result of the protocol data gets a high percentage of agreements from different coders. Third, EBD coding scheme is a dynamic and developing system. The author can add more factors for building more levels. Fourth, encoding design process can further understand the cognitive thinking of a designer. The next object is expected to quantify and improve the design process and to get more perfect and complete design results.
In the study, it not only refines EBD coding scheme, but also applies the scheme in the same design protocol. According to the data analysis, we can get some results. Although the lower level can get a high agreement and low variance, this level cannot grasp adequate information from the design protocol. The higher could get a large amount of information. Nevertheless the percentage of agreement is low, and variance of the third level is high.