Harry Potter revenue. The movie Harry Potter and the Sorcerer’s Stone opened as a great success. But every movie sees declining revenue over time. Here are the daily revenues for the movie during its first 17 days.

a) Without plotting the data, what components can you see in this series? Be specific.
For some series, a “seasonal” effect repeats weekly rather than annually. Here’s a regression model fit to these data with dummy variables for days of the week. (Day# counts days starting at 1.)
Dependent variable is: Earnings R squared = 96.9% R squared (adjusted) = 94.6%
s = 2.365 with 17 - 8 = 9 degrees of freedom Variable Coeff SE(Coeff) t-ratio P-value Intercept 21.0000 2.090 10.0 60.0001 Day# -1.42857 0.1194 -12.0 60.0001 Friday 12.4286 2.179 5.70 0.0003 Saturday 12.5238 2.166 5.78 0.0003 Sunday 7.95238 2.160 3.68 0.0051 Monday -4.28571 2.392 -1.79 0.1068 Tuesday -2.35714 2.377 -0.992 0.3473 Wednesday -0.928571 2.368 -0.392 0.7041

b) Interpret the coefficient of Day#.

c) Interpret the coefficient of Saturday in this model.

d) Predict what earnings probably were for Monday 12/3/01. What does this say about the model?

e) What probably happened to earnings after the initial 17 days?