test.diff.cor.single: Test for difference in correlation In AEBilgrau/correlateR: Fast correlations and covariances

Description

This functions uses the Fisher Z-transform (atanh) to test the null hypothesis of no difference in correlations between x1 and y1 versus x2 and y2.

Usage

 ```1 2 3``` ```## S3 method for class 'diff.cor.single' test(x1, y1, x2, y2, alternative = c("two.sided", "less", "greater"), conf.level = 0.95) ```

Arguments

 `x1` numeric vector, x-values for the first sample. `y1` numeric vector, y-values for the first sample. `x2` numeric vector, x-values for the second sample. `y2` numeric vector, y-values for the second sample.

Details

The `alternative` argument specifies the alternative hypothesis given below.

 H0: `cor(x1, y1) = cor(x2, y2)` `"two.sided"` => H1: `cor(x1, y1) != cor(x2, y2)` `"greater"` => H1: `cor(x1, y1) > cor(x2, y2)` `"less"` => H1: `cor(x1, y1) < cor(x2, y2)`

Value

A numeric vector giving correlation for each group, size-estimate and standard error, confidence intervals and p-values.

Author(s)

Anders Ellern Bilgrau <anders.ellern.bilgrau (at) gmail.com>

Similar usage to `cor.test` in `stats`, however not the same!
See `test.diff.cor` for a vectorised version.
 ``` 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19``` ```x1 <- rnorm(100) y1 <- rnorm(100) x2 <- rnorm(110) y2 <- 4*x2 + 0.5*rnorm(110) + 1 plot(x1, y1, xlim = range(x1, x2), ylim = range(y1, y2)) abline(lm(y1 ~ x1)) points(x2, y2, col = "red") abline(lm(y2 ~ x2), col = "red") diff.test <- correlateR:::test.diff.cor.single round(data.frame( two = diff.test(x1, y1, x2, y2, alternative = "two.sided"), les = diff.test(x1, y1, x2, y2, alternative = "less"), gre = diff.test(x1, y1, x2, y2, alternative = "greater")), 2) round(diff.test(x1, y1, x1, y1, alternative = "two.sided"), 3) round(diff.test(x1, y1, x1, y1, alternative = "less"), 3) round(diff.test(x1, y1, x1, y1, alternative = "less"), 3) ```