After the 9-21 Earthquake, some areas such as Hua-Shan Village of Gu-Keng Township, Yunlin County, were more easily affected by debris flow after serious storms. To protect these areas and promote rural development, the government has designed risk management and community development plans, spending hundreds of millions of dollars. This paper uses Hua-Shan Village as an example to illustrate the process of perceiving and evaluating the factors affecting debris flow disasters. We analyze the differences of perception and preference among respondents by using Likert scales, analysis hierarchy process (AHP), and multi-attribute utility theory (MAUT). We then compare these three methods according to the theoretical basis, simplicity, and empirical results for each one. It is found that MAUT combined with Likert scales is better than other methods for evaluating the perceptions and preferences based on theoretical reasoning, questionnaire design and field survey.