Analysis and Reanalysis of Shared Scientific Data

Abstract

Enormous strides in data handling and in telecommunications technology have made possible multiple access and sharing of data files. Sharing of research data often is resisted, however, because of (1) potential loss of monetary, political, or psychological reward; (2) potential conflict and disagreement; and (3) potential exposure of extreme bias or fraud. Found acceptable have been (1) the distribution of very large data files for further study, such as Public Use Tapes from the Bureau of the Census or the National Center for Health Statistics; and (2) the use of large data aggregates that serve as very comprehensive catalogs, such as bibliographic files for literature searches. Instances of scientist-to-scientist data sharing depend on individual arrangements. There are, however, many recorded instances of refusal to give access to data. Data that are liable to lead to conflict or controversy seem to be shared only if they fall within the realm of Freedom of Information legislation or court orders.