Correlation is a statistical tool that studies the relationship between variables. Correlation measure can take a range of values from -1 to 1. A correlation measure of 0 indicates that no relationship exists between any two variables. A correlation measure greater than 0 indicates that a positive relationship exists between two variables. A correlation value less than 0 means that a negative relationship exists between two variables. Correlation also indicates the strength of association between any two variables. A correlation measure that is closer to either -1 or 1 indicates a strong association exists between any two variables. If a statistical correlation between two variables doesn't make sense in real life, it is referred to as a spurious correlation. Spurious correlation is a relationship between variables that in real life have no logical connection yet the variables are related due to an unseen third occurrence ( Vigen, 2015).
Correlation 1.
The variables in correlation 1 are Math doctorates awarded and Uranium stored at US nuclear power plants (r = 0.952). A positive relationship exists between these two variables (0.952 is greater than 1). If Math doctorates awarded increases, Uranium stored at US power plants also increases. The association between the variables is also strong since 0.952 is closer to 1.
Delegate your assignment to our experts and they will do the rest.
The variables display the same pattern from 1996 to 2008 so they are related. From 2004 to 2008, the number of Math doctorates awarded increased from 1076 degrees to 1399 degrees. Within the same time period, the quantity of Uranium stored at US nuclear power plants also increased from 57.7 million pounds to 81.9 million pounds. From 1998 to 2000, the quantity of Uranium stored at US nuclear power plants decreased from 65.8 million pounds to 54.8 million pounds. Within the same time period, the number of Math doctorates awarded also declined from 1177 degrees to 1050 degrees. In addition, the correlation coefficient is 0.952. It shows that there is a strong positive relationship between Uranium stored at US nuclear power plants and Math doctorates awarded.
The relationship between Uranium stored in US nuclear power plants and Math doctorates awarded is spurious because the data set is completely unrelated. To allege that Math doctorates awarded is related to Uranium US power plants or vice versa would be to imply a spurious relationship. The variables have a mathematical correlation of 0.952 but have no sensible correlation in real life. This positive relationship may be as a result of pure coincidence or a third unseen factor referred to as common variable response.
Correlation 2
The variables in correlation 2 are divorce rate in Maine and Per capita consumption of margarine (r = 0.993). A positive relationship exists between these two variables (0.993 is greater than 0). If divorce rate in Maine increases, Per capita consumption of margarine also increases and vice versa.The association between the two variables is strong since 0.993 is closer to 1.
The variables display downward trending behavior from 2000 to 2009 hence they are related. The divorce rates were the lowest in 2009 at 4.1 per 1000. In the same year, the Per capita consumption of margarine was the lowest at 3.7lbs. In 2000, the divorce rates in Maine and Per capita consumption of margarine was the highest at 5 per 1000 and 8.2lbs respectively. In addition, the correlation coefficient is 0.993. It shows that a strong, positive relationship exists between Per capita consumption of margarine and divorce rates in Maine.
The relationship between divorce rates in Maine and Per capita consumption of margarine is spurious because the data set is completely unrelated. To allege that consumption of margarine causes divorces in Maine or vice versa would be to imply a spurious relationship between the variables. The variables have a mathematical correlation of 0.993 but have no sensible correlation in real life. This positive relationship may be as a result of pure coincidence or a third unseen factor referred to as common variable response.
References
Hill, R.C., Griffiths, W.E., & Lim, G. C. (2011). Principles of Econometrics.John Wiley & Sons, Inc.
Seddighi, H. (2012). Introductory Econometrics: A practical approach . New York, NY: Routledge.
Vigen, T. (2015). Spurious correlations .
McCallum, B. T. (2010). Is the Spurious Regression Problem Spurious? . Cambridge, Mass: National Bureau of Economic Research.
Ryan, T. P., & Wiley InterScience (Online service). (2009). Modern regression methods . Hoboken, N.J: Wiley.
