Can I get confidence intervals for Spearman's ranked correlations (or the difference between them)?? I've read contradictory reports from absolutely not, to treat them like Pearson's r, and do some kind of boot-strapping. Please don't tell me to normalise the data - the transformations render it meaningless.