! Develop valid earthquake casualty estimation models for planning and response (e.g., predisaster models of the medical impact of earthquakes for simulation of disaster impacts) (129,130).

! Determine whether and to what extent prior citizen safety training changes people's behavior under stress. Few studies have evaluated the extent to which people know about or follow earthquake safety recommendations or the effectiveness of recommended procedures (e.g., "duck and cover"). More research clearly needs to be done in these areas, and more effective teaching and training tools need to be developed.