New insights on goodness-of-fit tests for ranked set samples