Document Type


Degree Name

Master of Arts (MA)


Geography & Environmental Studies


Faculty of Arts

First Advisor

Barry Boots

Advisor Role

Thesis Supervisor


Previous assessments of factor analytic invariance to scale and aggregation effects have led to discrepant results. To determine the true effects, this study comprehensively examines the influence of scale and aggregation on factorial ecologies. This investigation is completed for three data sets, four scales, and thirty aggregations at each scale. Of these three data sets, two are artificial. These two data sets differ only by levels of spatial autocorrelations as one data set contains independent areal unit observations while the other set includes modest positive spatial autocorrelations. The third data set consists of variables from the 1986 Saskatoon enumeration areas. Several prominent themes emerge from these findings. when areal unit observations are independent, scale effects are trivial and aggregation effects are substantial" However, introduction of positive spatial autocorrelations among variables generates sizable scale effects and reduced aggregation effects. The theoretical data results are also moderately predictable from basic spatial unit data characteristics. Empirical results display considerable scale effects and modest aggregation effects. When increasing scale with the empirical data, communalities, eigenvalues, percentage of explainable data set variation, factor scores, and factor loadings are altered. These exact variations include increasing explanatory power of factor models with fewer significant factors and increasing generality of the largest unrotated factors. These findings along with several other modifiable results, attest to the substantial effects of scale and aggregation on factorial ecologies, with modifiable results from factorial ecologies, one must question the completion of contemporary factorial ecologies in geography.

Convocation Year


Convocation Season