The OP is asking about interpretation for Statistic value derived from cor.test function in R for Spearman correlation test. My doubt is the same, but for Pearson correlation test. Based on R material stat.ethz.ch/R-manual/R-devel/library/stats/html/cor.test.html these methods are different, but I am not sure if different enough to make another question. What do you think?
Also, I was not sure how to ask for this help (chat, meta, or just asking and wait if question is marked as duplicate). Thanks a lot.
@AndreSilva I do not understand the question at all, because its description of the referenced materials does not appear to match what is in those materials and it uses some terminology strangely. For instance, it distinguishes "the value of the statistical test" from both its p-value and its statistic (rho). What, then, is the "value"?
Since you appear to have understood what this question is asking, would you be able to explain it to me here?
@whuber. I will try: If you look at "values" for cor.test function in R, look for three of them: statistic (this is the one I would like to ask what it does mean), p-value (I think this is the probability of null hypothesis being rejected on the test. null probability is that x and y variables do not have collinear (Pearson) or monotonic (Spearman) relationship. The estimate (like the OP wrote: "rho") it is the number (between -1 and 1) that expresses the relationship between x and y ....
(e.g. if the estimate is 1, that means x and y has perfectly positive collinear relationship. In Pearson case, both variables x and y must be independent an assume Normal (Gaussian)distribution. I know this hipothesis test is based on t distribution (I think the Statistic value) has do with it, but not sure what exactly it means. If you did not understand, probably I did not too. Is that it? Thank's
@whuber "If method is "pearson", the test statistic is based on Pearson's product moment correlation coefficient cor(x, y) and follows a t distribution with length(x)-2 degrees of freedom if the samples follow independent normal distributions. If there are at least 4 complete pairs of observation, an asymptotic confidence interval is given based on Fisher's Z transform."
I was going to ask what this means. I did not get it. But, nevermind (this should not be so difficult to get it). I will study more. In case, I have progress I may try to post an answer on above question. Thank's for your attention.