(This continues exercise 1; it is still a hypothetical.) After thinking things over, Dr. Smith still believes that the response schedule is linear:

Yi,x = a + bx + δi, the δi being IID N (0, σ2). But she decides that her values for

a, b, and σ2 are unrealistic. (They probably are.) She wants to estimate these parameters from data.

(a) Does she need to do an experiment, or can she get by with an observational study? (The latter would be much easier to do.)

(b) If she can use observational data, what else would she have to assume, beyond the response schedule?

(c) And, how would she estimate the parameters from the observational data?