Crosbyfan said:
Cause and effect do not have a purpose.
Cause and effect do not care whether you can predict it.
However, a cause and effect study do... which is essentially the point we are after, and in fact pertains to this discussion. If you choose to continue to ignore that, that is up to you, but I won't take part in it anymore.
Smoking can cause lung cancer in certain people. Yes, but that certainly has nothing to do with a cause-effect study, which is what we are trying to prove.
We are trying to determine if something is predictive of the future here, which is a cause-effect study. Smoking and lung cancer are not predictive... the incidences for and against make the prediction scope way too wide. There are many more factors involved that prevent it from being a true cause-effect relationship.
Lack of knowledge and statistical data alone is not enough to prove cause and effect.
Equally: Lack of knowledge and statistical data alone is not enough to prove cause and effect does not exist.
Who is trying to prove anything with a lack of knoweledge? Obviously you'd need to know the basis of the test and relationship. I mean simply saying that 100% of the people in the study who drank orange juice got a hangover afterwards, without the other information of the study is essentially useless, but no one has done that yet.
No one is implying a lack of information.
It is very likely that eating peanut butter causes death in some people under some cicumstances. (compare that to my original statement on smoking and coalmining in post #260)
Requiring 100% statistical correlation is rubbish.
Granted, 100% correlation is next to impossible. But things like getting lung cancer from smoking, or dying from eating peanut butter aren't even 20% correlation. When you smoke, you do not necessarily get lung cancer. Smoking 2 packs a day for 40 years I can predict that you will get some form of ailment. It may not be lung cancer, but it may be heat disease, or high blood pressure. That is predictive. Smoking causes maladies for those who smoke for a long time. That is cause and effect. Virtually everyone who smokes for a long period of time will get some sort of ailment from it.
If a meteor lands on your house and kills you I would say there was a strong possibility it caused your death, even after I found out that one landed on your neighbours house, went through your neighbours brain and caused him permanent brain damage but he survived.
That's not true cause and effect though, at least nothing that you could apply for the purpose of predictive measures.
Think about it... why go through the purpose of a cause and effect study if you can't use the information from it for anything, other than looking back? Seems kind of useless doesn't it? Hence the reason why they say smoking increases your risk of lung cancer... because that is predictive. 96% of people with lung cancer were smokers. That is cause and effect. Smoking is the cause, and increasing your risk of lung cancer is the effect. The purpose isn't to find out why... it's to predict further results.
Similarly 100% statistical correlation does not prove cause and effect. (see orange juice)
Well naturally, when you run a fundamentally flawed study (like your orange juice one), 100% statistical correlation means nothing. Apparantly I gave you far more credit than you deserved... but I guess that is my fault. Cause and effect also requires other possible outlying conditions to be excluded first. How can you say it's orange juice when you haven't eliminated the other outliers first?