A Generalized Index of Dissimilarity

Abstract

The index of dissimilarity can be interpreted as the ratio of the number that must be moved from cells of excess to cells of deficit to achieve even distribution. This interpretation is used to generalize the index in two directions. First, the index is made applicable to more than two groups at a time. Second, an index and a test of significance are made available for explorations of cells of a two-way contingency table. DISSIM is the name of a computer program which provides these calculations for contingency tables.

Tolerancia a la diversidad y segregación residencial. Una adaptación del modelo de segregación de Schelling con tres grupos sociales

“…Si bien existen propuestas de generalización de este índice (Sakoda, 1981;Martori, Hoberg & Surinach, 2006), ellas presentan características que las alejan del índice original. Por ejemplo, el trabajo de Sakoda (1981) propone una generalización del índice de disimilitud, pero su propuesta no se basa en una generalización multidimensional, sino en una generalización que permite obtener de forma agregada el nivel de segregación de N grupos poblacionales.…”

“…Por ejemplo, el trabajo de Sakoda (1981) propone una generalización del índice de disimilitud, pero su propuesta no se basa en una generalización multidimensional, sino en una generalización que permite obtener de forma agregada el nivel de segregación de N grupos poblacionales. En consecuencia, aunque la interpretación del índice sigue siendo la misma que la de Duncan, el cálculo del índice propuesto por Sakoda es un poco engorroso, puesto que exige calcular para cada grupo de comparación su propio índice y luego sumarlos todos para obtener una medida agregada de segregación.…”

Global and local spatial indices of urban segregation

“…The generalized spatial dissimilarity index D m ð Þ is a spatial version of the generalized dissimilarity index D(m) developed by Sakoda (1981). The D(m) index is a measure of how population proportions of each areal unit differ, on average, from the population composition of the whole study area.…”

“…In the 1970s, segregation studies started to focus on multigroup issues, including the segregation among social classes or among White, Blacks, and Hispanics. To meet these needs, a second generation of segregation indices was proposed by generalizing versions of existing two-group measures (Morgan 1975, Sakoda 1981, Jargowsky 1996, Reardon and Firebaugh 2002. However, these measures are insensitive to the spatial arrangement of population among areal units.…”

“…Wong (1998a) proposed a spatial version of the generalized dissimilarity index D(m) developed by Sakoda (1981). In its original version, the D(m) index is a multigroup variant of the dissimilarity index D. Wong replaced the population counts of the tracts in the generalized dissimilarity index D(m) by composite population counts, which are obtained by grouping individuals that interact across tract boundaries.…”