Resumo:
Although segregation and accessibility studies has advanced in recent years and academia and
public planners have demonstrated considerable interest, even so these themes are treated as
isolated form and for the most part in big cities and metropolises. The location of Social Interest
Housing (HIS) is a big social problem in Brazil since they are allocated in places without urban
infrastructure and it is remoted from public services and equipment. In front of this context, this
work aims to analyze the relationship between socio-spatial segregation and accessibility in the
municipality of Itajubá/MG it is going to emphasize Social Interest Housing. To perform this
search, it was done a case study based on geospatial analysis in a Geographic Information
System (GIS) and statistical techniques. Firstly, it was examined socio-spatial segregation and
accessibility, where the result is the average of equal values obtained for each of the variables
considered. The Moran Global and Local Indexes were used to test the spatial autocorrelation
of two analyses, finally the Pearson Correlation Coefficient was applied to analyze the
correlation between them. The results showed us that lower segregation values originate from
some census areas in Centro and Pinheirinho districts, followed by other sectors around central
area. Peripheral areas have higher segregation value, where there are exactly two social
residentials units analyzed are found. With respect to accessibility, there is an opposite situation
due to central region has higher values for the detriment of peripheral regions. The Vitória
residential is situated in sector with higher socio-spatial segregation value and lower
accessibility value, conversely the Esperança Residential is in second class with higher
segregation values and in second class with lower accessibility values. It was pointed out that
results of two analyzes have considerable spatial autocorrelation, in other words, there is a
meaningful tendency for low or high segregation sectors to being contiguous to other with the
same characteristic, and low or high accessibility sectors to being neighbors to others with the
same results. Furthermore, they have a significant negative correlation with each other in 73%
of the sectors, in other words, higher value of socio-spatial segregation in a sector, lower is
going to be accessibility value for that sector, and vice-versa. The research presents practical,
scientific and social implications and it was contributed to studies of socio-spatial segregation
and accessibility based on multiple criteria in medium-sized cities.