Question

2. Jack and Rose attend a lecture about the contributing factors for interoceanic passenger ship accidents. The lecture shows regression results on severity of the accident (Y) and water temperatures (X). Rose believes that, because most ships travel through certain routes, temperatures (X) is highly correlated with other attributes of the routes, such as piracy and port availability. This makes it hard to really assess temperature's effect on the severity of an accident (Y). Jack's concern is that ships face more variation of severity of accidents in colder waters than in warmer waters, which makes results of a regression difficult to trust. Captain Smith hears this discussion and it is completely puzzled by it. He does not understand why these issues would make the regression result unreliable. Write a clear explanation for the captain about the two econometric issues discussed by Jack and Rose.

Answer 1

Rose concern is highly correlated with the other attributes i.e it is multicollinearity in data. And Jack concern is that ship face more variation i.e heteroscedasticity. These are the two econometrics issues discussed by Jack and Rose which would make the results unreliable.