风险寻求中的图形表征效应

Effects of Graphical Displays on Risk Seeking

在负性风险信息沟通中,采用文本表征与图形表征,结果发现图形表征方式更能引起风险回避行为。基于此,本研究以获益方案作为风险沟通信息,分别采用文本表征与图形表征来探索信息表征方式对风险寻求的影响。结果显示:在文本表征方式中,个体决策主要受备择方案之间风险大小的影响;而图形表征方式中,个体不仅受备择方案之间风险大小的影响,而且更受备择方案之间获益大小的影响。结果说明:相对文本表征,图形表征方式更能引起决策者为了获益而冒更大的风险,从而表现出风险寻求偏好,从而证实"图形效应"在风险决策中存在普遍性,为我们风险决策信息提供依据。

The influence of communication styles of risk information on decision-making is an important topic. One of the most important aspects of decision-making for individuals depends largely on how information on risk is presented to them. Recently the research has mainly focused on comparing the numerical displays with graphical displays. Concerning the effects of numerical and graphical displays on negative risk information, the research results found that graphical formats are more effective for increasing risk avoidance behavior and verified the existence of the graphical effects. These graphical effects are based on the situation of negative or loss events, but in reality, many of our decisions require us to take risks in order to achieve our goals, showing risk seeking preferences. The purpose of this research is to determine whether graphical formats are more effective for increasing risk-seeking behavior from the perspective of the gain information of risk Communication. Using numerical and graphical displays respectively for two investment items with different risks and benefits, we examined how individuals showed risk seeking in two different types of information representation. In this study, which employed a between-participants experiment design, participants were randomly assigned to one of the numerical and graphical display groups. Participants were then required to choose one of the two investment items to explore how the information display formats influence risk-seeking behavior. Four cognitive assessment tasks were used so participants would make a favored degree evaluation, ranging from favored （1） to very favored （7）, used to determine the mechanisms underlying the graphical effect. The software package E-prime 1.0 was used for materials presentation and data collection, and SPSS 13.0 was used for data analysis. Data results showed participants＇ risk seeking behavior preference in the numerical displays group and graphical displays group accounted for 52.94% and 68.78% of the total number, respectively. There were significant differences between them （X 2＂~＇4.50, p〈.05 ）; relative to numerical displays, more participants in the graphical displays group chose the item when benefits are substantial despite high risk, which means the graphical displays cause more participants to increase the preference of risk seeking. Thus, through data conversion, the results show that in numerical formats an individual＇s decision-making behavior was influenced by the magnitude of the risk between alternatives （the multiple correlation coefficient was .475, the multiple coefficient of determination was .225）. This was determined after merging the preference degree of two items into a new variable. The risk seeking preference index, as a criterion variable, and the four aspects of cognitive evaluation tasks, as prediction variables, also helped to make the multiple regression analysis. For graphical formats, individuals decision-making behavior was assessed under the influence of the magnitude of risk between alternatives （the multiple correlation coefficient was .432 and the multiple coefficient of determination was. 187） and under the size of profit between alternatives （the multiple correlation coefficient was .623 and the multiple coefficient of determination was .388）. The results show that the graphical representation can cause more risk seeking preferences under the benefit condition. In addition to previous research, these results verify that the graphical effects （1） Iniversally contribute to risk communication based on the characteristics of information, （2） Use different information representations to extract varying information, （3） Reach communication effectiveness, （4） Improve decision-making level, and （5） Avoid poor decision making.