http://rockvillecentral.com/2009/11/contributor-opinion-by-roald-schrack-2009-rockville-election-analysis.html/ News and views on Rockville, Maryland since 2007 Mon, 28 Feb 2011 21:01:44 +0000 hourly 1 http://wordpress.org/?v=3.3.2 http://rockvillecentral.com/2009/11/contributor-opinion-by-roald-schrack-2009-rockville-election-analysis.html/comment-page-1/#comment-2341 Erik Read Thu, 12 Nov 2009 13:36:28 +0000 http://rockvillecentral.com/?p=3937#comment-2341 Roald,Thanks for the analyses. I find myself surprised at your comment that there was no "slating". I calculated the correlation coefficients, covariance and even R-squared values for each candidate vote percentage across the Rockville districts relative to ex-Mayor Hoffman's vote percentage. I found that the candidates fell into three groups. Those which correlated with Hoffman (Britton, Gajewski, Moore), those which negatively correlated with Hoffman (Martin, Newton, Onley, Ovase) and those which did not correlate (Henn, Pierzchala). Further comparisons revealed that members within the first two categories tended to correlate more with each other and anti-correlate with members of the other group.Granted, the correlation coefficients are lower than those reported after the last instance of actual slate formation, but the pattern seems clear enough. So I am wondering what is your cut off for calling something a slate, even a qualified one? What do you think is the possible genesis of the pattern I have discerned? Roald,Thanks for the analyses. I find myself surprised at your comment that there was no “slating”. I calculated the correlation coefficients, covariance and even R-squared values for each candidate vote percentage across the Rockville districts relative to ex-Mayor Hoffman’s vote percentage. I found that the candidates fell into three groups. Those which correlated with Hoffman (Britton, Gajewski, Moore), those which negatively correlated with Hoffman (Martin, Newton, Onley, Ovase) and those which did not correlate (Henn, Pierzchala). Further comparisons revealed that members within the first two categories tended to correlate more with each other and anti-correlate with members of the other group.Granted, the correlation coefficients are lower than those reported after the last instance of actual slate formation, but the pattern seems clear enough. So I am wondering what is your cut off for calling something a slate, even a qualified one? What do you think is the possible genesis of the pattern I have discerned?

]]>