社会与工程心理学研究室知识库

   本研究从信息论的角度出发，用计算信源每个符号平均信息量的公式，计算图形编码方式的交通标志的不确定性，来实现对交通标志的量化评价。根据一定的样本空间，元素越多，每个元素的概率越小，在某一群体空间，一个标志所具有的含义越多，标志不确定性越大的原则，预测信息量是评价标志优劣的有效指标。研究中用静态和动态两个实验，来分析标志信息量与标志认读加工之间的关系。结果证明在静态和动态情况下，均随标志信息量的增大，达到一定学习标准的学习次数、反应时和错误率都增加。动态下的反应时显著高于静态下的反应时，速度也是影响信息加工的因素，速度快则对标志的加工反应时短而正确率低，负荷任务的绩效也显著下降。本研究采用7项主观评价指标对标志进行评价，从语意表达方面证实信息量能很一致地预测标志的明确性和辨认性。并且得出了信息量低的，即作为交通标志可用性高的图标的物理特征:简洁一具体一醒目一易于联想一普通。
   随着计算机图形用户界面的发展，这种量化的评价方法也可用于计算机图标的设计评估中。

    Based on the process of obtaining information of traffic signs  with  generalized communication system model, a formulation of Average Information Content of Source was applied to compute the uncertainty of pictorial traffic signs. And the Information Content of traffic signs was looked as a quantitative evaluation index.
    The more elements in one sample, the less the possibility of each element. So, average information content relating to the possibility of occurrence of a given definition for a traffic sign can be used to predict design quality.
    A Static experiment and a dynamic experiment were performed to analyze the information process of traffic signs. The results demonstrated that the learning times of getting certain criterion reaction time and error rate to traffic signs were in direct ratio with their information content. Moreover, the reaction time in dynamic condition was significantly longer than that of in static condition. Velocity was also an independent variable in the process of recognizing traffic signs.The faster the speed of presenting traffic signs, the higher the error ratio and the worse performance of loaded tracking task.
    A rating scale with seven items was used to evaluate traffic signs subjectively. Information content  of traffic  signs  could  predict  the  meaningful  and  identifiable  in  the  aspect  of communicativeness. The physical characters such as conciseness, concreteness, eye-catching,associability and indistinctiveness of low information content traffc signs which had the best usable were obtained.
    The possibility of using the methods and findings of this study to evaluate icons for icon-based human-computer interface was discussed in the background of fast development of icon-based human-computer interface.