Cluster Analysis Results : Analysis

1033 Words Sep 8th, 2014 5 Pages
Cluster analysis results – overview
Through initial HCA application we identified and eliminated from our analysis two Sydney suburbs revealed through the analysis as unique ‘outliers’ (in socio-economic change over time, and ‘overseas migration churn’ respectively). Having eliminated the associated distortion, a k-means cluster analysis was undertaken for the remaining 175 disadvantaged areas, based on the four area groupings which emerged from our initial HCA results. This produced a more balanced assemblage of members within each category – see Table 2. As shown here, area Types 2 and 4 both encompassed the majority of disadvantaged suburbs (77%), and also contained most of the relevant population within them (89%). As suggested by Table 2 substantial contrasts in the incidence of disadvantaged suburbs of different kinds in the three cities were apparent, raising issues discussed in the next section.
Table 2 – Summary of Typology Distribution
Suburb typology category Sydney Melbourne Brisbane All No of suburbs Pop (000s) No of suburbs Pop (000s) No of suburbs Pop (000s) No of suburbs Pop (000s)
Type 1 13 49 - - 1 2 14 51
Type 2 48 534 25 388 - - 73 923
Type 3 13 68 2 17 11 43 26 128
Type 4 15 106 23 261 24 184 62 550 Total 89 757 50 666 36 229 175 1,652
Source: Authors’ calculations Table 3 – Summary of Variables by Typology Category
Dimension Variable Disadvantaged suburbs Rest of city** City total Type Summary definition Type 1 Type 2 Type 3 Type 4 All
A…
Open Document