Thesis (MA)--Stellenbosch University, 2003. / ENGLISH ABSTRACT: South Africa has a long history of divided towns and cities. The grave inefficiencies and inequalities that developed between the racial communities during these periods must now be redressed in post-apartheid South Africa by integrating and unifying the physical and social structures of the country's urban settlements. In spite of the positive general trends in race relations and attitudes towards residential integration, South African towns and cities generally remain hyper-segregated. This could be an indication that White attitudes pertain only to the principles of integration, but that they do not actually want to live in integrated neighbourhoods themselves. The aim of this study is to assess the influence of racial preference in the dismantling or continuation of segregation in the South African town of Vredenburg during the postapartheid era. This is done by determining the influence that the population group composition of a neighbourhood has on the desirability of living in that neighbourhood when accounting for varying levels of crime and neighbourhood deterioration. A factorial survey questionnaire was used to gather the data, which were then analysed by way of multiple regression analyses. The results of the analyses indicate that the sampled residents of Vredenburg are generally not influenced by the population group composition of the neighbourhood. However, the more unsafe the neighbourhood, the more litter that is strewn about, the lower the housing quality and the more unfriendly the neighbours, the less respondents liked the neighbourhood. The results also indicate that members of the upper socio-economic class are more critical of their neighbourhoods and tend to evaluate them according to stricter criteria than the lower socio-economic classes do. The findings suggest that the racial composition of a neighbourhood per se does not significantly affect the attitudes of Vredenburg's residents towards a neighbourhood. Rather, high levels of crime and residential environmental deterioration are the factors that strongly affect both White and non-White people's views of a neighbourhood. Higher levels of crime and environmental deterioration are commonly associated with the lower socio-economic class. In the case of Vredenburg, vast socio-economic differences exist between the White and non- White residents of the town. These differences are not likely to change considerably in the short term. The continuation of these class differences will most likely be the cause of continued segregation in Vredenburg. Keywords: Apartheid city, Centralisation, Concentration, Evenness, Exposure, Factorial survey, Hyper-segregation, Integration, Multiple regression analysis, Neighbourhood characteristics, Racial preference, Segregation, Segregation indices. / AFRIKAANSE OPSOMMING: Suid-Afrika het 'n lang geskiedenis van verdeeelde stedelike gebiede. Die erge ongelykhede en oneffektiewe strukture wat tussen die verskillende rassegroepe binne die stedelike gebiede ontstaan het, moet reggestel word in die post-apartheid era. Dit moet gedoen word deur die verdeelde fisiese en sosiale strukture van Suid-Afrika se stedelike gebiede te integreer. Ondanks die algemene positiewe neiging in rasseverhoudings en houdings teenoor residensiële integrasie, bly Suid-Afrikaanse stedelike gebiede steeds hiper-gesegregeerd. Dit kan 'n teken daarvan wees dat Blankes se ingesteldheid slegs positief is teenoor die beginsel van integrasie maar dat hulle nie self in geïntegreerde woonbuurte wil bly nie. Die doel van die studie is om die invloed van rassevoorkeur te bepaal in die aftakeling of voortsetting van segregasie in Vredenburg, Suid-Afrika, gedurende die post-apartheidsera. Dit word gedoen deur die invloed van bevolkingsgroepsamestelling op die begeerte om in daardie buurt te woon te bepaal, in ag genome die invloed van verskillende vlakke van misdaad en omgewingsverval binne daardie woonbuurt. 'n Faktoriale opnamevraelys is gebruik om data in te samel. Die data is daarna ontleed deur middel van veelvuldige regressie-analises. Die resultate van die analises toon dat die inwoners van Vredenburg, wie aan die steekproef deelgeneem het, in die algemeen nie beïnvloed is deur die bevolkingsgroepsamestelling van 'n woonbuurt nie. Daarteenoor het die deelnemers minder gehou van woonbuurte wat meer onveilig is, waarin meer rommel gestrooi is, waarvan die behuisingskwaliteit laer en die bure meer onvriendelik is. Die resultate toon ook dat lede van die hoë sosio-ekonomiese klas meer krities is oor woonbuurte en geneig is om dié areas volgens strenger kriteria te evalueer as die laer sosio-ekonomiese groepe. Die bevindings dui aan dat die rassesamestelling van 'n woonbuurt per se me die ingesteldheid van die dorp se inwoners beduidend beïnvloed nie. Dit is eerder hoë vlakke van misdaad en residensiële omgewingsverval wat beide Bruin en Blanke inwoners se opvattinge oor 'n buurt beduidend beïnvloed. Hoër vlakke van misdaad en omgewingsverval word gewoonlik met die laer SOSIOekonomiese klas geassosieer. In Vredenburg se geval bestaan daar groot sosio-ekonomiese verskille tussen die Blanke en nie-Blanke inwoners van die dorp. Dit is onwaarskynlik dat hierdie verskille in die korttermyn beduidend sal verander. Voortgesette klasverskille sal waarskynlik die oorsaak wees van volgehoue segregasie in Vredenburg. Trefwoorde: Apartheidstad, Blootstelling, Egaligheid, Faktoriale opname, Hiper-segregasie, Integrasie, Konsentrasie, Meervoudige regressie-analise, Rassevoorkeur, Segregasie, Segregasie- indekse, Sentralisasie, Woonbuurtkaraktereienskappe.