I came across a sutra in a data analysis course: Correlation is not causation!
Meaning: if X and Y variable change in the same way over a period of time, it does not prove that X causes Y (or Y causes X). It is quite possible that there is another hidden factor (Z) causing the changes in both X and Y.
For a long time, medical world believed that excessive alcohol consumption caused Lung cancer. That is what statistical studies showed. Alcoholics had a high number of Lung cancer. But a deeper study found a third factor: smoking. Most alcoholics were also heavy smokers!
But the statistical study itself can not prove that alcohol does not/ can not cause lung cancer. Who knows, it may have some role? But it stands to reason, that third factor - smoking- should be given its due importance. Why? because alcohol goes to stomach and smoke goes to lungs. Smoking delivers carcinogens directly to the lungs. Therefore, now we believe that it is more important to reduce smoking, to prevent lung cancer. A little wine is supposed to be good for health. Wait, where are you off to?
This is a good use of our common sense: looking at evidence, interpreting it wisely- rather than simply believing our subjective experiences: I have been doing X for last six months and I am stammering much less now, so it must be because of X only. In last six months many other things have been happening to me. What about those? I have done many many more things earlier- may be, they are delivering their results after some years now?
Most of us pride ourselves for being intelligent. Let us use it to analyse and learn from our experiences in an intelligent way. Let us also share our experiences to warn others.
Let me conclude on a lighter note: A study found that area of Texas reduced (it is next to Atlantic ocean) between 1910 to 1970 at the same rate as its population increased. A student concluded: with the increasing weight, land tilted and went under sea..really?!!
So, let us realize, that two changes may happen at the same time but may have NO relationship whatsoever. A little scepticism is not a bad thing after all.. Read more on this..
Meaning: if X and Y variable change in the same way over a period of time, it does not prove that X causes Y (or Y causes X). It is quite possible that there is another hidden factor (Z) causing the changes in both X and Y.
For a long time, medical world believed that excessive alcohol consumption caused Lung cancer. That is what statistical studies showed. Alcoholics had a high number of Lung cancer. But a deeper study found a third factor: smoking. Most alcoholics were also heavy smokers!
But the statistical study itself can not prove that alcohol does not/ can not cause lung cancer. Who knows, it may have some role? But it stands to reason, that third factor - smoking- should be given its due importance. Why? because alcohol goes to stomach and smoke goes to lungs. Smoking delivers carcinogens directly to the lungs. Therefore, now we believe that it is more important to reduce smoking, to prevent lung cancer. A little wine is supposed to be good for health. Wait, where are you off to?
This is a good use of our common sense: looking at evidence, interpreting it wisely- rather than simply believing our subjective experiences: I have been doing X for last six months and I am stammering much less now, so it must be because of X only. In last six months many other things have been happening to me. What about those? I have done many many more things earlier- may be, they are delivering their results after some years now?
Most of us pride ourselves for being intelligent. Let us use it to analyse and learn from our experiences in an intelligent way. Let us also share our experiences to warn others.
Let me conclude on a lighter note: A study found that area of Texas reduced (it is next to Atlantic ocean) between 1910 to 1970 at the same rate as its population increased. A student concluded: with the increasing weight, land tilted and went under sea..really?!!
So, let us realize, that two changes may happen at the same time but may have NO relationship whatsoever. A little scepticism is not a bad thing after all.. Read more on this..
No comments:
Post a Comment