The problem I'm having with this is that since the derivative of the e^X is just e^X and X is assumed to be a whole number as it didn't specify otherwise, this distribution should end up being the same for both Y and Y'.

Secondly, you tell us to assume that the conditional distribution of Y "satisfies the five attributes of classical regression." I'm not sure what you mean by "satisfies." It's already been stated that the distribution is symmetric so I couldn't say that the skewness attribute has been satisfied.

I may be looking at this problem all wrong, but I appreciate any input you may have.

Is anyone finding this textbook very hard to follow? The questions have very little to do with what we read and seem to use a whole different text book. Does anyone have a source they have been using outside the Fox textbook to help them understand this material better? Please help.

[NEAS: Most mathematical exam problems follow the formats in the practice problems on the discussion forum. The practice problems have much detail, so that you can follow the intution. The problems explain the material in the Fox textbook and have more detail, so that you can apply the methods to exam problems]