B Correlation vs Causation: What Can Be Inferred?

[gl0Wyrm]

If variable X is positively correlated with variable Y, under what circumstances am I allowed to infer that X does not causally prevent Y with reasonable probability?

I don't know much about statistics and I'm trying to motivate myself to learn by doing. For example: https://drive.google.com/open?id=0B9wG-PC9QbVEUTBIVmgxeElCQnc (https://drive.google.com/open?id=0B9wG-PC9QbVEZ2NEU3BvYzkwM0U)

 

[Ygggdrasil]

In general, you can't infer causality solely from correlations in observational studies. You would need to perform some randomized, controlled experiments in order to establish causality.

 

[chiro]

Hey gl0Wyrm.

If you want to assess causality then you should assess conditional probabilities in terms of a well ordered stochastic process.