The study of a disaster which has occurred cannot be researched using the traditional techniques as it is not possible to conceive or conduct controlled, randomized experiments for such an event. Paper I of this series described non-experimental, scientific methodologies which were applied to study the detection-extrication-resuscitation activities which occurred following the devastating 1988 earthquake in the Republic of Armenia, USSR. This paper critically evaluates the methodologies used for the Armenia study and proposes modifications in these methods for application to the study of future disasters.
Approaches which could be applied to the study of future disasters are defined and critically evaluated from the view of reliability, validity, costs, and practicability. The revised set of protocols is discussed in terms of: 1) the structured interview process; 2) training of personnel to conduct, synthesize, and evaluate the interviews; 3) the time required to complete the interview process; 4) sampling techniques; 5) mechanisms for cross-checking the data; and 6) the addition of preliminary data collection immediately following or during the event. Use of this revised approach should assist in the collection and analysis of data associated with future disasters so that it is possible to: 1) further enhance life-saving and reduce mortality; 2) improve relief efforts; 3) reduce damage to communities; 4) evaluate the long-term effects of such events; and 5) assist in better preparation for future events.