人机系统中广泛采用计算机合成语音为决策者提供建议。为了比较合成语音与自然语音以及直接和含蓄的沟通风格对不同国家决策者的影响,通过实验的方法,招募了24位中国学生和24位德国学生参与决策实验,测量决策者对建议的接受度,自信水平的改变度,感知的影响,以及对建议者的信任度、满意度和未来合作意向。统计分析结果显示,相比自然语音,合成语音建议显著提高决策者的自信水平,但是决策者对合成语音建议感知的影响较弱且未来合作意向较低。总体来看,直接的沟通风格比含蓄的沟通风格更容易被决策者接受并获得决策者更高的评价。当采用直接的沟通风格时,德国决策者比中国决策者感知到更强的影响;当采用含蓄的沟通风格时,中国决策者比德国决策者对建议的接受度更高。研究结果为合成语音软件的设计提供了指导。

Synthesized voice is widely used in a human-computer system to provide recommendations for decision makers. In order to compare the influence of the synthesized voice and the natural voice as well as the explicit and implicit communication styles on decision makers of different nationalities, a decision making experiment is conducted with 24 Chinese students and 24 German students. The acceptance of recommendations, the change of confidence level, the perceived influence, the extent of trust in the recommendation provider, the satisfaction degree, and the future collaboration intention with the project are identified in the experiment. The statistical results show that compared with the natural voice, the synthesized voice significantly increases the decision maker's confidence level, but the decision makers perceive lower influence and have less future collaboration intention under the synthesized voice. Overally, the decision makers accept explicit recommendations more often than implicit recommendations and have more positive evaluations on explicit advisors than implicit advisors. For explicit recommendations, German decision makers perceive more influence than Chinese decision makers; for implicit recommendations, Chinese decision makers accept recommendations more often than German decision makers. Based on the results, the recommendations for the synthesized voice software design are made.