Dataset highlighting importance of causation versus correlation


Hi All,

Thank you for all the really wonderful datasets you have out there.  My request is for a dataset (or two) in which drawing causal information would create incorrect conclusions.

In my college Stat 101 class, I recall looking at a dataset that compared Ice Cream Sales (independent variable) to Crime Statistics (dependent variable).  The data shows a strong correlation, but one should obviously be wary of drawing any causal relationships. Currently I just make up a data set to use in my current 8th grade class for this, but would love to use real data. Do you have any suggestions for other datasets where students could look for a correlation, be wary of the causation and then perhaps brainstorm outside variables (i.e. temperature outside) that impact both variables? Thank you,


5 years ago by greenawaltmath