Correlation and significance with dependent observations


New Member
I am trying to correlate two variables of which one variable has dependent observations as it is data derived from a social network. I am intending to use Spearman's rank correlation coefficient.
As I understand conventional methods for significance test do not provide reliable results. Resampling or bootstrapping might be alternatives. I am however not sure which method is best to use. And I could not find any literature on this issue. it would be great if somebody knows about useful literature.