In many cases, there are important metrics called the "strength of correlation". This is an important consideration when determining a causation. Additionally, there is the necessity of determining alternative causes. For example, when a school does better on some sort of testing after several teachers are fired, it COULD be because those teachers sucked that bad, or it could be directly related to the change in morale with the other teachers, or it could even be related to a change in management style, or a change in classroom size, or any number of other factors.
When one considers that a series of earthquakes are seen that correlate with fracking sites (biggest earthquakes ever recorded, always within 2miles of the site in multiple sites), there is precious little else to consider as likely alternatives other than a very unlikely set of happenstance or coincidence.
It's certainly possible that it's a coincidence, but a strong correlation tends to indicate that this is not the case. Understanding statistics at a deep level will ehlp you understand this more.
ALL surveys show a correlation. Inferring a causation is simply trying to eliminate as many other co-correlations as possible and demonstrating that the original correlation holds up even when other possible causes are removed.
Can you think of other causes for unusually strong earthquakes happening to cluster around fracking sites?
mad hatter azerbaijan ryan howard ps i love you ray charles cheney heart transplant weather san diego
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.