correlation is the relationship between two sets of data. as the english word implies, a high degree of correlation suggests that the data (and the events they represent) "move" in a similar fashion. but correlation does not imply causality: the rise in women's rights through the 20th century has been paralleled by the rise in obesity among americans. one cannot therefore conclude that increased women's rights leads to obesity, or vice versa! or perhaps one can: an increase in women's rights provides more independence for them to nurture their young, often to unhealthy levels of consumption of food, leading to obesity ;-).
there are a myriad of subtle ways in which random (or seemingly random) correlations can be investigated and/or explained. the potential confusion in the conclusion "the spread of the internet has led to increased belligerence among members of the left" is of a different nature from the women's rights <=> obesity idea above.
--ravi