OK, so this is 1970 to 2011, unfortunately 2012 only had 4 countries. SO it’s for the 15 European countries (fixed missing Germany before 1990.) Three sets of results are reported bellow for completeness: for both sexes combined, for males only and for females only. In each case there is regular Pearson coefficients, Pearson on logged death rates and Spearman coefficients.
Each set of results reports the four quadrants in addition to the four half-way splits and the total for the whole surface.
I also added a year by year graph of the correlations for each set (each cobmination of log/coefficient type). These seem a lot clearer than the quadrants I’d say, showing a dramatic change after 1993, though more so for females than for males. (Not sure what’s up with 2012 though.. )
Pearson on linear data:
## left right total
## top -0.13648 -0.5411 -0.4279
## bottom 0.06386 -0.6648 -0.6860
## total -0.20710 -0.6156 -0.4462
Pearson on logged data:
## left right total
## top -0.14487 -0.5798 -0.4892
## bottom 0.06376 -0.7069 -0.7378
## total -0.21968 -0.7282 -0.5822
Spearman
## left right total
## top -0.1532 -0.5379 -0.4483
## bottom 0.0704 -0.7375 -0.6233
## total -0.2003 -0.7967 -0.5595
Pearson on linear data:
## left right total
## top -0.04705 -0.4998 -0.3410
## bottom 0.14497 -0.5887 -0.6092
## total -0.14117 -0.5836 -0.4028
Pearson on logged data:
## left right total
## top -0.04475 -0.5340 -0.3925
## bottom 0.14579 -0.6192 -0.6470
## total -0.14901 -0.6815 -0.5212
Spearman
## left right total
## top -0.06092 -0.4774 -0.3600
## bottom 0.14977 -0.6578 -0.5216
## total -0.10836 -0.7236 -0.4893
Pearson on linear data:
## left right total
## top -0.2829 -0.5707 -0.5048
## bottom -0.1431 -0.6862 -0.6766
## total -0.2907 -0.6255 -0.4658
Pearson on logged data:
## left right total
## top -0.3130 -0.6129 -0.5758
## bottom -0.1597 -0.7450 -0.7610
## total -0.3197 -0.7517 -0.6177
Spearman
## left right total
## top -0.3030 -0.5849 -0.5486
## bottom -0.1600 -0.7739 -0.7218
## total -0.3392 -0.8421 -0.6128