A correlation coefficient measures the relation between two variables. The correlation coefficients in this study measured either the relation between elections or between votes in elections and background variables such as Catholicism. The basic unit of measurement was the percentage vote in a county for a given candidate. For example, if the relation between support for Roosevelt in 1936 and Stevenson in 1952 was being analyzed, each county would be assigned two scores--the percentage of votes cast for the Democratic presidential nominee in the two elections. The highest correlation coefficient possible is 1.00, which would be obtained if there was a parallel rate of increase in the percentages given the candidates in the two elections. For example, if the vote in County A was 45 percent for Roosevelt and 40 percent for Stevenson, in County B 50 percent and 45 percent, respectively, in County C 55 percent and 50 percent, County D 60 percent and 55 percent, and this pattern of one-to-one increase persisted in all the counties in the state, then a 1.00 correlation would be obtained. If, however, for County E the Roosevelt vote was 61 percent and the Stevenson vote was 85 percent, the parallel rate of increase would be disturbed. Even though for each county the higher the support for Roosevelt, the higher the support for Stevenson, the magnitude of the correlation coefficient would be below 1.00.