Correlation matrix. Statistical packages can easily calculate correlation coefficients for multiple pairings of variables. Results are often reported in the form of a correlation matrix. Figure 14.18 displays the correlation matrix for data from a study of geographic variation in cancer rates.r The variables are:

CIG cigarettes sold per capita BLAD bladder cancer deaths per 100,000 LUNG lung cancer deaths per 100,000 KID kidney cancer deaths per 100,000 LEUK leukemia cancer deaths per 100,000 Notice that the values for each correlation coefficient appear twice in the matrix, each time the variables intersect in either row or column order. For example, the value r = 0.704 occurs for CIG and BLAD and for BLAD and CIG. The correlation of 1 across the diagonal reflects the trivial fact that each variable is perfectly correlated with itself.

Review this correlation matrix and interpret its results.