Overall, there was a positive correlation between the mean score of the risk perception of the target hazard and the weighted mean by the risk perception of the relevant hazards. The positive correlations were found for each hazard too in both years. The only exception was an earthquake. In 2004, as you can see, an earthquake has a great disparity. This is because an earthquake has been recognized that the risk perception is very high but the relevance to other hazards is low. However, the Japan earthquake made aware of the impacts of an earthquake on various hazards, especially nuclear power generation or water damage. After all, the gap between the expected and the actual means in an earthquake is narrowed, we think. So we consider our hypothesis was supported.