Design and Analysis of Reliability Studies—the Statistical Evaluation of Measurement Errors
Design and Analysis of Reliability Studies—the Statistical Evaluation of Measurement Errors
复制标题
DOI：
10.1136/jech.45.2.175
复制
发表时间：
1991-06
影响因子：
6.3
通讯作者：
E. Arnold
中科院分区：
文献类型：
--
作者：
E. Arnold
Longitudinal Study, a study of the role of birthweight in areal variations in infant mortality, and increased attention to the mortality of migrants, and variations according to place of birth. The examination ofmortality from selected causes back to 1921 shows how long some of the variations between areas have persisted. It is perhaps a pity that the authors have not paid more attention to the influence of age in variation between areas. The use of the SMR as a summary measure implies that the ratio of death rates in two populations is constant over age strata, or at least not sufficiently variable as to make the SMR misleading. It is interesting that the regional variation in "all cause" mortality is rather wider for the age groups 15-64 than for 65 and over, but this finding is only briefly commented on. Local authority areas were used as the main units of analysis, although it was recognised that they are internally heterogeneous with regard to their social, economic, and environmental characteristics. Such heterogeneity inevitably limits investigators' ability to explain the variations in mortality that are observed. The suggestion that future volumes may make more use of smaller areas such as wards is therefore welcome. Such minor criticisms do not detract seriously from the value ofthis volume, which will be of interest to epidemiologists, public health specialists, and anyone interested in geographical variations in health.
