| Summary: | 博士 === 國立臺灣科技大學 === 設計研究所 === 96 === Some graphics can be easily distinguished and desired, while some cannot be desired or comprehended by the general public. What are the causes behind such outcomes? Relevant investigations in great deal have been found in the gestalt theory, design discipline, ergonomics, and object recognition theory. The purposes of this study are centered on understanding the means of measuring whether a graphic can be easily recognized and proposing a “Good Graphic” design. The three critical factors presented for the study in this field are: (1) Revelation of a method of measuring the most optimal graphic simplification value and, (2) establishment of the basis of quantitative description for graphic design (3) provision of reference for the clinical graphic simplification design.
As the preliminary stage, this study has adopted the literature review and expert interview to conduct the survey on the current state. In addition to the compilation of relevant visual theories, this study has deduced the following graphic simplification design methods. The methods are mainly categorized into two models: “Extraction of the Overall Conformation” and “Extraction of Partial Features”. The former can be further divided into the sketch of outlines, preservation of structural relationship, and stylization. The later can be divided into the emphasis on visual features, emphasis on functional features, preservation of texture, and material characteristics.
The second stage of the study attempted to construct the frame of the quantitative graphic description through the discipline of experiment in combination with the clinical design operation. By utilizing the regression analysis, the quantitative description of the external graphic shapes and the assessment of the degree of internal simplification cognition would yield the formula defined by the graphic simplification operation as: Graphic Simplification Degree＝1.276＋0.005(Overall Node Quantity)+0.131(Partial Element Quantity). This stage, in addition to the proposition of the simplicity design operation and measurability for the “Extraction of the Overall Conformation” and “Extraction of Partial Features,” graphics generated from the operation rules can also be used as the samples for the next experiment stage. Such would improve the limitations found in the manipulation of the experimental graphic samples in the previous studies for rendering the statistical results much more accurate and the analysis of design rules much more clarified.
The third stage was characterized by the tri-factor experiment design to investigate the effect of the three independent variables – simplification combination, object category, and participant profile- on the visual recognition. Based on the evaluation results generated from the three dependent variables – recognition accuracy, reaction time, and preference, and then “the most optimal graphic simplification values” could be comprehended. Analysis results of the ANOVA for total scores indicated that the “graphic recognition accuracy” was significantly correlated with the main effects of the simplification combination and the object category, and their interaction. The reaction time was individually influenced by the simplification combination and the participant profile at a significant degree while being simultaneously impacted by the interactions between the simplification combination and the object category. The graphic preference was significantly impacted by the main effect of the simplification combination and was simultaneously influenced by the interactions between the simplification combination and the participant profile.
After the conduction of the study at various stages, a number of findings were achieved. First of all, the graphic details were conducive to facilitate the promotion of the graphic recognition accuracy and reaction time; however, when the detail quantity is reaching a certain threshold value, the graphic recognition accuracy and reaction time could no longer be enhanced, even with an increase in the graphic details. For individuals with the design background, the degree of preference might be even decreased. Fundamentally, when the overall graphic node quantity exceeds 25 or more or the partial element quantity exceeds 3 or more, a faster reaction time and an approximately 80% accuracy can be achieved for the graphic recognition. We would recommend the designers pay exceptional attention to the effects exerted by the object category on the recognition accuracy and reaction time when simplifying the design, especially the simplification model for the elimination of unsuitable nature object parts. Second, even though the designers and non-designers exhibit no variations in the recognition accuracy, differential preferences over the degree of graphic simplification are discovered. Designers’ preference and degree of simplification present an inverted U curve relationship, while non-designers exhibit the positive linear correlation. Designers must exercise due attention to this variation.
The operation of object graphic designs in this study can be illustrated from two perspectives. From the perspective of overall form, the Attneave Graphic Quantification Theory can be applied to compute the number of nodes. From the perspective of the partial elements, the geons presented by the RBC Theory can be used for calculating the partial element quantity. A number of simple computation formulas deduced by the study can predict the evaluation values of the graphic recognition accuracy, reaction time, preference, and cognition of simplification degree. The scope of the graphic simplification design method application is comprehensive, including the trademark, public signs, style design, and various computer-related interface and illustration designs. Designers frequently process the graphics based on sensuality and intuition. The study has proposed clear design operation guidelines and numerical values while not attempting to replace the originality of the design. It is, however, intended to compensate the insufficiency in the designer’s intuition or to enable the design beginners with the rules to follow. Through these prediction methods, references and recommendations can then be provided to the designers when assessing the graphics. Moreover, the study is desired to be established as the fundamentals for the quantitative descriptions of the computer-aided graphic applications and various graphic design methods in the future.
|
|---|
