|
The Influence of Errors on the Correlation of Ratios
F. E. A. Briggs
Abstract
If correlations are calculated between ratios containing a common variable the presence of observation errors for the common variable may lead to spurious inferences about the correlations obtaining in the population. By making assumptions regarding the distribution of the observation errors it is possible to adjust for the influence of the errors on the correlations of the ratios. The adjustments are derived on various assumptions and are applied to correlations in which this problem originally arose [4].
|
