In: Economics
Review the above pdf document on "Spurious Correlations" and then write a paper that argues why correlation does not mean causality.
Page limit: maximum 2 pages. Calibri or Arial 11 font, single-spaced, margins (1”) on all sides, no covers or binders, just a plain cover sheet with your name. Unlimited exhibits, graphs and attachments as long as they are referred to directly in your text. Cover sheet, and attachments not counted in page limit.
Sad Sate of Research These Days Spurious Correla6ons
(Do not have acess to PDF, points made from general debate).
Correlation: Linear relationship between two variables.
Correlation checks for the linear relationship between any two variables, i.e., whether the variables move in the same direction, opposite direction or have no relationship.
Often correlation is taken to mean causation. But such an inference is incorrect. Correlation simply sees how the two variables are moving in relation to one another. For instance if height and weight are postively correlated, we can expect the weight of an individual to be higher if the height is higher and vice versa.
HOWEVER, it would be wrong to imply that higher height causes higher weight or that higher weight leads to more height.
Correlations can often be suprious, i.e., two variables can have a significant correlation without any apparent cause or reason. This would just mean that the variables move together/against each other. For instance, we have a positive correlation between length of an individual's hair and their ability to jump. That doesn't mean that because an individual's hair is longer, they can jump more.
Causation means one variable leads to another. A low birth weight can result in severe health problems in the later life. Causation is determined from theory or our basic understanding of the world. It requires knowledge of the subject.
We have seen earlier than correlation does not imply causation. Similarly, just because one variable may cause another, it doesn't mean that their correlation will be high. Why? Because correlation measures only linear relationship. Causation need not be linear causation.
Therefore, we must be very careful while interpreting the numbers in front of us.