2015年5月20日(水)13:30より6号館7階6701号室にて、University of Sydney の Nicholas Irving Fisher先生， Korea University の Jae Chang Lee 先生による講演がございます。講演タイトルと要旨は以下の通りです。

Title

Some Problems related to Statistical Data on Disaster

Speaker

Jae Chang Lee (Korea University)

Abstract

Reliable statistics on disasters are essential for informed policy decision making as well as scientific endeavors to reduce their risks. Statistical standards related to disasters in general are not conceptually well defined. Definitions and adequate conceptualizations must be established before other attempts for better communications and meaningful scientific analysis. Reliability and precision of disaster statistics depend heavily on internationally agreed definitions and protocols for collecting related data. Many historical disaster statistics are dubious and not very reliable. Many of disaster statistical data are based on numbers mainly coming from local governments and various independent agencies.

The United Nations' Fundamental Principles of Official Statistics (1994) can serve as a check list for these statistics. Also the International Statistical Institute's Declaration on Professional Ethics (2010) is a valuable source for this purpose.

Accuracy of statistics should be maintained properly to meet the intended use of them and also follow the rules for scientific measurements. These will be discussed in terms of disaster statistics with a few historical examples.

Title

Some Roles for Statistics in Avoiding Disasters

Speaker

Nicholas Irving Fisher (University of Sydney & ValueMetrics Australia)

Abstract

Disasters may be natural and man-made, avoidable and unavoidable. There is little that can be done to prevent a natural catastrophic event such as an earthquake, volcanic eruption, cyclone or tsunami from wreaking great havoc. When cyclone Tracey destroyed Darwin on Christmas Day, 1974, the consequences in terms of loss of life and damage to buildings and property were unavoidable.

Natural catastrophic events do not necessarily have to result in disasters: a variety of precautions can be taken to reduce the impact on life and damage to property and the environment. Thus, the disastrous outcome when the Fukushima Daiichi nuclear power plan was hit by a tsunami on 11 March 2011 was found, in the resulting enquiry, to be man- made, with the consequences of such an event foreseeable.

On the other hand, man-made disasters such as the 2008 Global Financial Crisis, or the sinking of the Korean ferry Sewol on 15 April 2014, or almost every workplace incident resulting in death or injury to person or equipment are avoidable.

Statistics has a vital role to play in all of these situations. Much technical work has been directed into forecasting natural disasters. In the area of man-made disasters, Statistics has long played a role in managing risk, for example through actuarial calculation. However, there are important situations where Statistics has contributed rather less than it might, perhaps because there is rather less technical challenge. It is some of these areas that are the focus for this talk: how can Statistics help avoid another Global Financial Crisis, another ferry disaster, and more workplace deaths and injuries?