Scale Changes and Impacts on Factorial Ecology Structures

Only in the past few years has any attention been directed toward oblique rotations and higher-order analysis in factor analytic research. This paper focuses upon the utility of higher-order analysis by comparing factorial-ecology results for a single set of variables over two differing systems of observational units covering the same area. The necessity for higher-order analysis is demonstrated for sub communities, but not for census tracts. Interpretations of underlying structural configurations from higher-order analysis lead to different perspectives than from lower-order findings. Implications of the results are then related to issues of areal aggregation and structural change.