How, in general, do you determine whether X causes Y or Y causes X?

How, in general, do you determine whether X causes Y or Y causes X?



Answer: You run a cross-lagged study. You measure both variables at two (or more) points of time, using the same subjects at both times so that you now have (at least) two pairs of X & Y for every subject. If X is the cause of Y, then the correlation between X-at- time-1 and Y-at-time-2 (hereafter: X1-Y2) will be stronger than the correlation between Y-at-time-1 and X-at-time-2 (Y1-X2). On the other hand, if Y is the cause of X, then the Y1-X2 correlation will be stronger than the X1-Y2 correlation. In general, because we think that causation always happens forward in time, the cause should be measured before the effect to get the best correlation.

Comments

Popular posts from this blog

Examine the collapse of Han. What factors explain how this occurred? How influential in Han Dynasty on Chinese History?

JAD Advantages and Disadvantages

Summarize effective communication strategies in organizational hierarchies, and review the role and relevance of the organizational grapevine