Let the User Beware

Two reports in this issue converge with findings which, in essence, constitute a warning to researchers. Professors Jack Vincent and Llewellyn Howell have agreed that it is unwise to place great trust in the World Event Interaction Survey (WEIS) and the Conflict and Peace Data Bank (COPDAB). Both scholars have carried out series of data analyses, using first one event data collection and then the other, and both have found that the results are importantly different. Neither Howell nor Vincent are ready, at this time, to recommend the use of one data set and the avoidance of the other. Both make references to probable practices of 'over-collecting' and 'under-collecting', but their investigations to date have not provided detail sufficient to pinpoint evidences of faulty workmanship in the data acquisition process. Equally for foreign policy making and international theory, Vincent and Howell have sounded an alarm with regard to the probable validity and reliability of the two major sets of international event data. They are not the first to suspect event data, however. Objections have been voiced and doubts raised, with persistence and vigor, from the first appearance of event studies. Almost any group of critics can be relied upon to react with 'standard' queries, among them: (1) Won't event files, being based on daily public reporting of world news, be biased seriously because of the selection and suppression of stories by reporters, editors, and owners of the mass media?; (2) Since heads of state, foreign ministers, intelligence chiefs, and military leaders do much of their work in foreign affairs in secret, how can event data collections be regarded as valid-how can they reflect the real circumstances of world affairs, in the face of so much secret activity?;