THE SAMPLE CLUSTER: A NEGLECTED DATA SOURCE

In: Public opinion quarterly: journal of the American Association for Public Opinion Research, Band 34, Heft 4, S. 593-603

Abstract

R's in large-scale cross-sectional survey samples are generally clustered into areas. 3 analytical tools for exploiting the sample cluster for substantave analysis are illustrated: (1) The intra-class r coefficient (rho) is used as a measure of homogeneity of characteristics within clusters. (2) A comparison of the degree of homogeneity across types of questions indicates some sources of agreement. (3) The use of the sample cluster as the smallest unit of analysis explores other sources of agreement about the existence of a neighborhood org, the block club. The data are drawn from an area sample of Detroit Negroes (N=619) from 122 clusters of 3 to 8 D. U.'s. The homogeneity of opinions on several issues & beliefs about the existence of neighborhood problems & block clubs are found not to be greatly affected by demographic characteristics of R's. AA.