Which of the 'following is NOT a guideline for establishing causality? Perform a randomized, controlled experiment: Look for cases where correlation exists between the variables of a scatterplot Take into consideration all the other possible causes_ Keep all variables the same to get duplicate results.