Cause and Effect

Third, the correlation technique produces only the probability that the corresponding happenings are not random (or fortuitous), not that one phenomenon either causes the second to happen or is a precursor (early warning signal) of the second. A significant correlation coefficient only indicates the degree to which more research may be worthwhile. Often it is found that the two correlated events are on different branches of a common tree. For example, it may be that whenever certain types of clouds appear on the western horizon there is a significant correlation with the number of admissions to city hospitals; and more of those special clouds mean more admissions. It would be silly to suppose that the clouds cause people to go to hospitals or that the hospital admissions cause the special clouds. Further study would probably show that the special clouds are indicative of imminent severe thunderstorms and that the increased hospital admissions are from car accidents related to poor visibility and slippery roads in the ensuing bad weather.

0 0

Post a comment