Statistics Help?

Question

Consider a multiple regression of Y (gas mileage) on two predictor variables X(1) (vehicle weight) and X(2) (engine type). Suppose X(2)is an indicator variable that takes the value +1, if the engine is fueled by gasoline and 0,if the engine is fueled diesel. Is there some numeric value for variable X(1) such that if X(1) is less than this value the mean fuel economy for gasoline vehicles will be lower than the mean fuel economy of the diesel vehicles, but if X(1) is greater than this value,the mean fuel economy for gasoline vehicles will be higher than the mean fuel economy of the diesel vehicles.

Yes, there is such a value.
No, no such value exists.
The question depends on the sign of the slope parameter for variable X(1)
You can only answer this question for a specific regression on a specific data set.

Answer 1

I'm confused. Why not use the actual vehicle weight for X1? This would be a simple multiple regression problem with two variables-- one variable is the actual weight and one is an indicator variable for engine type. Then you could determine which variable is more important in determining the response variable (gas mileage).
Why make it so complicated?

Answer 2

I would answer the last option,"you can only answer this question for a specific regression on a specific data set".

Comparing fuel efficiencies based only on X(1), the vehicle weight, would not be a very comprehensive or useful analysis.