Using data from a review of 225 samples studied after 132 disasters, I aimed to determine (a) if the methods in disaster research have been adequate to capture the complexity of disasters, (b) if the methods influenced results observed, (c) if there has been progress toward use of superior methods, and (c) if there are methodological innovations emerging in other fields that would serve this field well. Most samples (72%) were assessed only once after the disaster, and very few (4.4%) were assessed before the disaster. Over half (61%) were assessed within 6 months of the event, a sizable percentage (28%) within 2 months of the event. Most often, samples were selected on the basis of convenience (31%) or because they formed a census of an affected group (27%), with random (19%), purposive (17%), and clinical (6%) samples being used less often. Sample sizes varied widely, from 11 to over 5,000, with a median of 150. These characteristics differed across samples according to sample type (youth, adult survivor, rescue/recovery), disaster location (USA, other developed country, developing country), and disaster type (natural, technological, mass violence). Controlling for these characteristics, samples that were assessed before the disaster, selected for reasons of convenience, or large in size tended to show less severe effects than other samples. Certain desirable study characteristics (e.g., longitudinal designs, representative samples) have been decreasing in prevalence over time, while others (early first assessment) have been increasing. Innovations such as latent trajectory modeling or hierarchical linear modeling might advance the field’s ability to capture the complexity of disasters, but the field still needs to attend to the fundamentals of sound epidemiologic research. (Review posted to www.redmh.org March 2005.)

Download Methodology paper here (PDF document- 830KB)