What do we wish to do ? We want to get good F ratios in ANOVA for the pairs of variables and find the more discriminant ones with Discriminant analysis. A linear correlation table will help. All data is given in the table. We have a good pictorial or geometric sense and can concentrate furiously on the first 2 charts. Birth rate (TFR) is the most discriminatory variable among variables. In Mexico, rural south has 4 TFR, whereas the industrial north has 2TFR.