著者
大山 正 中原 淳一
出版者
公益社団法人 日本心理学会
雑誌
心理学研究 (ISSN:00215236)
巻号頁・発行日
vol.31, no.1, pp.35-48, 1960 (Released:2010-07-16)
参考文献数
26
被引用文献数
15 19

If a cross figure consists of upper and lower black arms, right and left white arms, and a central gray square, we do not perceive the mosaic sum of these five parts. We see, instead, a black vertical bar and a white horizontal bar. In the central part of this figure, we can see simultaneously two colors, white and black, one behind the other. One of these two colors appears transparent and the other appears to be seen through the former (See Fig. 1A). The following investigation is concerned with this apparent transparency.Method: Experimental procedures were nearly the same with those of our previous studies on figure-ground reversal (This journal, 1955, 26, 178-188). Observers were instructed to fixate their eyes upon the center of the figure 60 to 120 seconds, and push the first one of the three electric buttons when the black bar appeared in front of the white bar, push the second button when the white bar appeared nearer, and push the third button during ambiguous appearances. As the measure of relative dominancy of black and white, the formula, Rb=100Tb/(Tb+Tw), was adopted, in which Tb indicates the total pushing time of the first button, and Tw, that of the second.Results: 1) In general, the white bar has a stronger tendency to appear in front of the black bar when the central square is light gray, and the black bar is dominant when the central square is dark gray. It was discovered that the relative dominancy was approximately proportional to the difference between the square root of reflectance of the central part and that of the arms (Table 1, Fig. 2, 3, 4).2) The lightness of the surrounding field has little effect on the relative dominancy of two bars (Table 2).3) When the arms are of two of the four chromatic colors, red, yellow, green and blue, instead of white and black, and the central part is the mixture of these two colors produced by the rotating disk, yellow is the most dominant color, red is the second, green the third and blue the last. However, red may be more dominant than yellow if the above mentioned effect of lightness is eliminated (Table 3).4) When the vertical arms are red, the horizontal arms are green, and the central square is the mixture of red and green in various ratios, the relative dominancy is represented in a S-shaped curve as a function of the mixture ratio, i.e., the angle ratio in color disk (Table 4, Fig. 5, 6).5) The effect of area of the arms is equivocal. There are large individual differences, and the difference of instructions easily affects the results (Fig. 8, 9, 10, Table 5). The similar results were obtained in the stimulus figures of another type (Fig. 1B, 11).
著者
中原 淳一
出版者
帯広畜産大学
雑誌
帯広畜産大学学術研究報告. 第II部, 人文・社会科学篇 (ISSN:03857735)
巻号頁・発行日
vol.4, no.4, pp.217-225, 1976-02-25

Taxonomy of 2×2 games has been shown by Rapoport & Guyer. Later, Hamburger introduced a metric classification system of 2×2 games restricting his examination on separability of payoffs. If the classification system is a metrical one, then not only comparability of game behavior of strategicaly different games, but also quantitative analysis of game behavior is supposed to be possible by using the parameters of the system. Moreover, dynamic game methods should become a powerful experimental method for the study of interpersonal interaction processes, if the system contains metricaly related, psychologicaly meaningful games such as prisoner's dilemma game, chicken game and so on. Following the above preliminary considerations the author presented a new way of construction of the 2×2 game system. Itemized discussions are as follows : 1. A state vector is attributed for each player. The element of this vector is a potential payoff. A rectangular arrangement of these vectors makes a state matrix. The state matrix of a two-person game is shown as follow : [numerical formula] 2. Somewhat ad hoc payoff rules are applied to the state matrix, and the 2×2 payoff matrix is constructed as follow : [numerical formula] 3. Characteristics of games which are deducible from this parametric payoff matrix are discussed. 4. A symmetric case (α=β, x=y) is examined at first, and α is hypothesized to be a fixed parameter. In this case, game are quasi-chicken games if x>α, prisoner's dilemma game if α>x>α/2, quasi-coordination games if α/2>x>o, and pure coordination game if x=0. 5. An asymmetric case (a=β, x≠y) is considered next. In this case, if y>α>x, then column player's payoffs are always secured as positive, and also he can determine row player's payoff as positive or negative only by his own strategic choice. This game is called the absolute positive-negative control game. 6. Finally, several further extensions of the method are discussed.