Author Question: What is spurious correlation? Provide a real or hypothetical example where two variables might ... (Read 88 times)

kfurse · **on:** Jun 24, 2018

What is spurious correlation? Provide a real or hypothetical example where two variables might exhibit such a relationship.

Question 2

What is residual analysis, and what information can it provide?

jasonq · **Reply #1 on:** Jun 24, 2018

Answer to Question 1

Spurious correlation occurs whenever two variables are highly correlated but there is little evidence to suggest any causation. For instance, the closing of the Dow Jones Industrial Average (DJIA) could be highly correlated with the winner of the Super Bowl. (When the NFC wins it has historically tended to close higher for the year, and when the AFC wins it has tended to close lower for the year). However, there is no evidence to suggest that the Super Bowl winning conference causes the DJIA to move up or down. Likewise, there is no evidence to suggest that the movement in the DJIA causes one conference or the other to win the Super Bowl.

Answer to Question 2

Residual analysis refers to a broad class of tests which can be performed using the error terms from a regression line to test the assumptions of regression analysis. Residual analysis can test the assumption of normal distribution of the error terms, the assumption of equal standard deviations of the error terms, and the assumption that the error terms are independent from one another. In residual analysis, we may also flag error values that are especially large.