Dissertation/Thesis Abstract

Classifying Changes in Neighborhood Racial Composition: St. Louis City and St. Louis County, 1970-2010
by Jones, Matthew, M.S., Southern Illinois University at Edwardsville, 2018, 169; 10844022
Abstract (Summary)

This thesis set out to classify changes in the neighborhood white-black racial composition for St. Louis City and St. Louis County Missouri from 1970 to 2010. Relying solely on decennial census data as accurate depictions of neighborhood change may not be the best solution, mainly due to the inability to capture neighborhoods undergoing transitional stages between decades. To remedy this problem, methodologies were developed and implemented to better depict these transitional stages temporally.

Using the Longitudinal Tract Database (LTDB), census tracts were used to depict St. Louis City and St. Louis County neighborhoods, all while using 2010 tract delineations to allow temporal analysis to occur. All 305 census tracts and their changing white-black racial compositions were individually analyzed, leading to the development of eight different neighborhood typologies: majority white and black enclaves, gradual black and other transition, white flight and delayed white flight, white re-entry, and multi-racial integration neighborhoods.

After testing for spatial autocorrelation, two rounds of K-means cluster analysis were conducted on the rescaled white, black, and other percentage variables from 1970 to 2010 with the use of various spatial weight matrices (i.e., Rook, Queen, and distances of 4, 6, and 8 miles). The resulting clusters were analyzed further to identify the same neighborhood transition typologies that were identified in the pre-spatial statistics approach. The second tier K-means clustering on the spatially rescaled attribute variables produced neighborhood transition typologies that eliminated spatial contiguity issues from the first K-means clustering approach. Results from all K-means clustering analyses were consolidated, producing one final depiction of neighborhood transition typologies for St. Louis City and St. Louis County.

Latent class mixed models were conducted on both white and black groups. The resulting analysis produced six neighborhood transition typologies that were further consolidated into one final depiction of neighborhood transition typologies. The consolidated neighborhood typologies from the K-means clustering analyses and the latent class mixed models both depict similar spatial and temporal patterns of neighborhood transition typologies.

Indexing (document details)
Advisor: Hanlon, James
Commitee: Brown, Stacey, Zhou, Bin
School: Southern Illinois University at Edwardsville
Department: Geographical Studies
School Location: United States -- Illinois
Source: MAI 58/01M(E), Masters Abstracts International
Subjects: Geography
Keywords: K-means cluster analysis, Latent class mixed models, Longitudinal tract database, Neighborhood transition, Neighborhood typologies, Spatially rescaled variables
Publication Number: 10844022
ISBN: 978-0-438-37897-1
Copyright © 2020 ProQuest LLC. All rights reserved. Terms and Conditions Privacy Policy Cookie Policy