In the four-year project, we plan to combine the psychophysics method and systems factorial technology to design experiments, analyze data, and make inferences. We examine the processes in multi-attribute decision making, including the process architecture, decisional stopping rule, and processing capacity. In addition, we examine the decision strategy is influenced by various contextual factors. The results enable us to verify and modify our previously proposed hypothesis – relative salience hypothesis, to discriminate various types of decision models (compensatory decision strategy vs. non-compensatory decision strategy), and to examine whether decision process is involved the emergent feature. The purposes are as follows: (1) to investigate how decision makers integrate the information from multiple sources to make a correct and efficient decision, (2) to investigate how a decision maker integrate his own decision evidence and the advisor’s decision information, and (3) to investigate in the context of group decision-making, how cooperation and competition affect the information integration process. We design a novel probability inference task where participants are required to learn the predictability of two features and utilize the learned probability to make a categorization decision. In the first year, we manipulate the feature combination (integral feature and separable feature) and relative feature predictability to test the multi-retribute decision process, In the second year, we manipulate the reliability and confidence of an AI advisor to test how the two factors affect the human-machine joint decision. In the third year, we test the effect of competition and cooperation between two decision makers on the information integration process in joint decision-making. In the fourth year, we create a novel virtual reality game to test all the manipulations mentioned in the previous three years in order to increase the ecological validity of the present study. Our project has several theoretical and practical significance. Our results can shed light on the decision models and which factors may affect the decision-making process, which also highlight the flexibility of the decision strategies. Moreover, our results can provide a guideline for design of the human-computer interaction and a better cooperation way for joint decision-making while making decisions under a uncertain context.
|Effective start/end date||21-08-01 → 22-07-31|