Exam 2 with Part 1 Instructions

March 2, 2015

Exam 2 information will be posted here. Check back for details.

• Part 1 (take-home) details will be be announced in class on Wed/Thu Mar 4-5.
• Part 2 (in-class) is scheduled for Wed/Thu Mar 11-12

Part 1 (take-home) Instructions:

Download Pueblo Real Estate Sales Data: Exam 2 Part 1 (MW 11am – the Google Sheet file was corrupt. Please use this file instead.)

There are 2 worksheets:

• 2014 SF Sales (1,418 records)
• 2015 YTD SF Sales (166 records)

Like Exam 1, the last digit of your PID determines which column of data you will use for the exam. If your PID ends in 3 your assigned column is “PID 3”. If your PID ends in 4 your assigned column is “PID 4”. And so on. Your assigned column has either a 1 or a 0 in each row. You will include each row listed with a 1 in your assigned column. You can ignore rows with a 0.

Use your assigned “2014 SF Sales” data to build a regression model (or a set of regression models) designed to accurately forecast “Selling Price”.

Use your assigned “2015 YTD SF Sales” data to test your regression model(s). Generate forecasts for each assigned row and compare forecast vs actual selling price in two ways:

1. Create a scatter plot with “Actual” on the horizontal (x) axis and “Forecast” on the vertical (y) axis.
2. Calculate MAD and MSE for your set of forecasts.

Deliverables:

1. A set of descriptive statistics and graphs summarizing the 2014 data.
2. Regression Output for your forecasting model(s).
3. Scatter plot comparing Actual to Forecast selling prices.
4. MAD and MSE calculations.
5. 1-page “executive summary” describing your modeling process and results.

Think of this as a consulting project. You have been asked to develop a statistical model to forecast Single Family property sales in Pueblo. Part 1 is the initial report you will present to your client.