| i. Suppose that Starbucks tends to have a sale on cold brew when the temperature is lower. Does this suggest Cov(T,P1) is positive, negative, or 0?i. Suppose that Starbucks tends to have a sale on cold brew when the temperature is lower. Does this suggest Cov(T,P1) is positive, negative, or 0? ii. Do you think an increase in P1 would increase or decrease QCB, holding all other factors fixed?iii. Do you think ˆ β1 is a consistent estimator of ˜ β1? If not, using your previous answers, do you think it is systematically above or below ˜ β1 in a large sample? ( pick above or below.) iii. Do you think ˆ β1 is a consistent estimator of ˜ β1? If not, using your previous answers, do you think it is systematically above or below ˜ β1 in a large sample? ( pick above or below.) iv. Suppose the analyst has a richer dataset, now with prices. Specifically, the analyst has a large sample of independent and identically distributed observations of QCB,T, and P1 for each day. Can the analyst use data on prices to construct a better estimator of ˜ β1 than the previous estimator, ˆ β1? If so, describe such an estimator. (You would regress what on what?) (You may assume there are no other factors that systematically vary with T and P1. More formally, you may assume MLR.4 holds.) |
