Search

In many scientific investigations, it is desired to bring together, or link, two or more documents which represent the same individual, even though these documents do not contain a unique identifier and were derived from different sources. In medical and public health research and elsewhere, this problem is known as the document linkage problem. This paper considers some aspects of classifying pairs of documents into one of two populations when their items are identifying information, where each item of information can take on three distinct values correct, incorrect or missing. Section I identifies three document linkage problems. Sections 2 and 3 deal with the mathematical formulation of the multivariate document linkage problem. Section 4 gives the classification procedure and Section 5 deals with the application of the theory to a problem in the field of public health. [ABSTRACT FROM AUTHOR]

Data are available to study the variation in mortality in this country occurring month by month for fully six decades. Death rates by month have been computed and published for the expanding death-registration area of the United States of 1900-1932 and for the complete nation (48 States and ultimately 50 States) from 1933 to the present. Even a cursory glance at the available data leads to the obvious conclusion that a definite pattern of seasonal variation of deaths exists. This seasonal swing is characterized by a high death rate in the early part of the year dropping to a trough in the summer months, and rising again during the latter part of the year. The general pattern is well recognized, as are similar seasonal curves for various causes of death. Despite the apparent wealth of statistical material at hand relatively little analytical matter has been produced in this country measuring seasonality of mortality as compared with the wide range of data compiled for statistical series portraying the nation's economic operations. Yet seasonal variation is receiving increasing attention in the evaluation of statistics of vital events. The very factors which make seasonal indices so valuable to economists and analysts also have proved to be of service to demographers and public health researchers. Seasonal adjustments provide refined information about current trends; they separate the seasonal components of statistical data from cyclical and irregular fluctuations; they provide indices with which to measure the significance of month to month change. Seasonal variation in the crude birth rate has been analyzed for the period 1933-40 using the link relative method [1]. Seasonal indices for 1948-54 were prepared with the aid of Univac, the high-speed electronic computer of the United States Bureau of the Census, using an adaptation of the standard ratio to moving average method [2]. The indices prepared by the Bureau of the Census became the basis for the regular computation of seasonally adjusted figures of births by the National Vital Statistics Division, U. S. Public Health Service 131. Similarly, marriage data published by the National Vital Statistics Division have been "deseasonalized" with the aid of methods developed by the Bureau of the Census [4]. This paper presents the results of an effort made with the assistance of Univac to eliminate seasonal variation from mortality data. Some of the particular problems encountered in adjusting mortality data, and the attempts to surmount these are described. [ABSTRACT FROM AUTHOR]