JAIST Repository: 未経験者による機器操作アイコン設計の観察的研究

全文

(1)JAIST Repository https://dspace.jaist.ac.jp/. Title. 未経験者による機器操作アイコン設計の観察的研究. Author(s). 真下, 奈緒子. Citation Issue Date. 2003-03. Type. Thesis or Dissertation. Text version. author. URL. http://hdl.handle.net/10119/449. Rights Description. Supervisor:下嶋篤, 知識科学研究科, 修士. Japan Advanced Institute of Science and Technology.

(2) Designing icons for machine operations: an observational study of the case of inexperienced designers Naoko Mashimo School of Knowledge Science Japan Advanced Institute of Science and Technology March 2003 Keywords : icon design, abstraction, inexperienced designer Most functions of common hardwares, such as microwaves, laundry machines, and VCRs, are displayed by characters. Although displays of functions of software programs are often in pictorial displays, designing effective icons generally requires a special skill (Horton 1994). The purpose of this study is to identify what exact problems one encounters in designing pictorial displays of machine functions. In this study, we first examined textual and pictorial displays of functions in common hardware and software interface, sorting them by Horton's classification scheme (Horton, 1994). We observed that particular difficulty arises in expressing "action" functions in Horton's sense, and many of those operations are displayed vaguely by the depictions of their final states and other typical states associated with them. As the technology develops, our software and hardware appliances will be equipped with novel functions, and designing effective pictorial displays of them will be a challenge even to talented and experienced designers. Also, the Copyright © 2003 by Naoko Mashimo. 1.

(3) development of internet environment will give opportunities for even non-experts to design pictorial displays of operation buttons. It is therefore important to analyze typical problems that non-experts encounter in designing displays of action functions. As a pilot study, we conducted a preliminary observation of the way people design. icons for machine functions, and the way people use those icons. We. found that the displays thus designed had many design elements that users never have recognized. We hypothesized that the created displays will be affected by whether creators perceive themselves as designers or simply as users. We designed our main observation to test this hypothesis. Our experimental system was the software program that can perform 4 different operations on 25 blocks in the screen. The blocks are different in color (red, blue, and yellow) and in shape (circle, square, and triangle). The four operations are all action functions in Horton's sense, and they are (1) sorting the blocks by color, (2) sorting them by shape, (3) enlarging each block, (4) shrinking each block. The total of 20 subjects participated, of which 4 subjects were assigned the designer role and the remaining 16 subjects were assigned the user role. The designer subjects were further divided to two groups. One group went through the following procedure: they were first (1) given a design direction, (2) asked to operate the system, (3) asked to describe the function that they thought the system was performing, and then (4) they designed symbols. The other group followed the same steps but in a different order, namely (2)-(3)-(1)-(4). Contrary to our hypothesis, we observed no significant difference in displays designed by. two groups of subjects: whether or not subjects were in the. designer's position had little effect on the number of design elements. Next, we analyzed the way the user subjects conceived the displays designed by the designer subjects.. We found that user conceptions are affected by the. methods that the designers used to "abstract" unnecessary features of target operations. For example, we can abstract color properties either by removing all colors from the displays, or by including all colors in them. This way, those who. 2.

(4) look at the symbols can not take specific color as the object of operation. We could see two patterns in which users were misled by different abstraction methods. 1. Certain particular matches of the three colors and the three shapes were displayed in the symbol, although the designer succeeded in including all colors and shape. 2. A particular shape (either circle, square, or triangle) remained in the symbol, although the designer succeeded in removing all other shapes. Number 1 was a prominent case where the display shows imperfect abstraction. There were a number of cases where a display accidentally showed a particular shape, although the designer clearly tried to abstract shape from his/her display by excluding all particular shapes. Users tended to misinterpret displays in those cases. In contrast, misinterpretation was decreased when the designer abstracted shape by including all shapes in his/her design. Number 1 is another case of imperfect abstraction. The blocks operated in the system come with three colors and three shapes. Thus, there are nine possible matches of colors and shape. Number 1 above is the case where the display showed only specific matches from nine possible matches. It is considered that this partial abstraction led users to misconceive the displayed operations. Displays that show nine patterns would not allow the user to see any specific matches as the operational object. We performed χ2 -test in user responses to displays with 9 matches and those with 3 matches. We found a significant difference in the frequency of user misinterpretation in these two cases. These results suggest that the imperfect abstraction is a major cause of user misinterpretation of operational icons. When we change the linguistic expression to the graphical expression, we need to display it as a visible symbol. In this process, various accidental features may be represented in the design, although the designer mean to abstract them. This will give chances of varying interpretations to those who look at the expression. Since the view of graphical expression is left to individual arbitrary selection, its information which a. 3.

(5) designer did not expect may increase（Shimojima, 1996）. Moreover, when we try to express the information "gathering red figures" by putting in the upper part of the display, even if it is accidental, it can be interpreted to mean "gathering red figures in the upper part". This is related to "accidental feature" and it means that it may be impossible to express certain weak information with a picture. This phenomenon was also observed for displays that. performed abstraction which take in all colors and all form. Although these. symbols included all nine matches of color and shape, they displayed each group of blocks in line. As the result, some users misinterpreted the operation as gathering blocks in line. We found that these factors make it difficult to pictorially display action functions. We hope our finding will provide a useful guidance to those who design symbols of the ACTION function.. 4.

(6)