3. Your company produces Halloween-themed figurines made from silica which are targeted to amateur radio enthusiasts ("Ham Sand Witch"). You've collected data on annual sales (S, number of figurines sold per year) the price of your figurine (P, in dollars), and the population size in the different cities in which your company has retail locations (N, number of people; you currently only have one establishment in each city). You estimate the following regression model: S = a + bP + cN. In your regressions, you usually look for a 10%-or-better level of confidence.

a. What signs do you expect for a, b, and c?

b. Your regression yields the following results:

Adjusted R Square	0.853
Independent Variables	Coefficients	Standard Error	t Stat	P-value
Intercept	6177	1192	5.183	0.00352
P	-193.26	29.71	-6.504	0.00128
N	0.0065	0.0041	1.607	0.16893

Interpret what these coefficients mean.

c. Does price have a statistically significant effect on sales?

d. Does population have a statistically significant effect on sales?

e. What portion of the total variation in sales remains unexplained?

f. Ham Sand Witch is considering selling figurines in a new city, where the population is 310,000, and setting its figurine price at $32. What level of annual sales would you expect in this new city (rounded to the nearest unit)?