Regression Analysis - positive correlation


I am running a regression analysis using data for aircraft values. The problem that I have encountered is that GDP global and Oil Price are heavily positively correlated. However, all things equal, rising Oil price has a negative effect on most aircraft.

Any ideas how I can include Oil price in the analysis so that it has the correct sign (negative)? If I do it in one go Oil simply is shown as having a small positive effect due to correlation with GDP...
Why would you include two variables if one can do the job? If they are both heavy correlated, they both explain (roughly) the same thing... just choose one and stick to that one.
Because from business point of view as well as for model accuracy I need the two factors to be taken into consideration. I am not saying that both need to be included in regression analysis but I need to add the effect of the second one, possibly in later step.

Just looking for ways to filter that out and process my regression line (achieved from analysis with GDP only) to improve the thing.
Last edited:

I would do a Hierarchical multiple regression with the "independent" variable first (Oil Price) and the dependent latter (GDP) to see if inclusion is relevant.
Thanks Clbustos,

In the last decade it was rather the GDP that has driven the oil prices and not the other way round so I would have to turn the sequence.

I'll try that one but if anyone has more idea please do contribute!