Controlled experiments can only establish the existence of a correlation between two events and not a causal relationship.

Controlled experiments are an attempt to show that some correlation is in fact a causal relationship. In such experiments, two sets of data are collected. One set (the experimental group) is allowed to be influenced by the potential causal factor, while the other set (the control group) is not. If the effect shows up for the experimental group and not the control group, then it is likely that the correlation being studied is a causal relationship.