I would like to begin my remarks by taking issue with the approach to regional phenomena in Canada implied in Professor Irvine's paper, “Assessing Regional Effects in Data Analysis.” Irvine deals with the problems of locating and quantifying statistically significant (and meaningful) relationships among behavioural data categorized by geographical regions. The implicit assumption in this approach is that the existence of regions is not in question, only the extent of their relevance to the interpretation of political data.