1. PS2_data_2016.xlsx is an Excel file that contains housing transaction records of a

town for 76 quarters (19 years). Each observation corresponds to a transaction,

and one house can appear in multiple observations. The dataset has 22 variables,

and brief variable description is provided in the worksheet “Info.” The data also

contains 76 dummy variables (d_q1 through d_q76) indicating the quarter of

sales for each house.

1. Calculate mean, median, standard deviation, minimum, maximum,

percentiles at 1%, 5%, 10%, 25%, 75%, 90%, 95%, 99% of “Living Area” and

“Price”.

1. Calculate the percentage of “YES” response for all the dummy variables.
2. Estimate a hedonic regression of log house price on all the variables in the

dataset using the data in “Hedonic Data” worksheet.

1. Using estimated coefficients on quarterly dummy variables, construct a

house price index (with 1st quarter as base quarter) and draw a graph

(quarter on the horizontal axis and index on the vertical axis).

1. “Repeat Sales Data 3” worksheet only includes transaction information for

houses sold more than once. Compare characteristics of the whole dataset

and the repeat sample

1. Estimate a repeat sales regression of log house price using the data in

“Repeat Sales Data 3” worksheet. Assume all the housing characteristics

didn’t change between sales.

1. Repeat the part (d) using regression results from the part (f).
2. Estimate a repeat sales regression of log house price using the data in

“Repeat Sales Data 3” worksheet again. However, do NOT assume all the

housing characteristics didn’t change between sales, but include all the (first

order differenced) housing characteristics in the regression.

1. Repeat the part (d) using regression results from the part (h).
2. Compare three housing price indices ((d), (g), (i)) and explain why the

constructed indices are not the same.