If your boss gave you data from a survey that showed the IQ of each person and the IQ of his or her mother and then asked you to create a report, the first step would be to plot the data on an x and y-axis.The x variable and y variable may be interchangeable in this circumstance because the person who was surveyed more than likely their IQ is affected by the IQ of their mother.It would be reasonable to think that these two variables have some type of connection, but according to the concept of correlation and causation, one variable is not causing the outcomes of the other variable even if there is a strong correlation.In this case, if the x variable was the IQ of each person interviewed that would be plotted along with the y variable which would be the IQ of his or her mother.Once these variables are plotted on a scatterplot, it would be easier to see if they have a positive or negative slope.Another way of seeing the relationship is by entering the data into a simple linear regression model and seeing how the slope fits into the data points.If it is within the data points and there is a positive slope, it would be safe to say that as the IQ of the surveyor increases so does the IQ of his or her mother.If the data points are not in line and kind of all over the place, this would be a non-linear data model and there would be other ways of handling that information.

It seems subjective, but what is a good test to know if data has some sort of correlation and causation?