BUS 3700: Project (Regression analysis)This is to be done individually. Important- submit the report as a Word document.
1. Collect data on about ten to fifteen single family houses for sale in one locality (in one ZIP code) on these
variables (using websites such as realtor.com or zillow.com or redfin.com):
i. address
ii. area in square feet,
iii. number of bedrooms,
iv. age in years, and
v. asking price.
Try to get a random sample: for example, do not select all high priced or low priced houses; perhaps you could
take say every third item in the listing. Make sure that all houses that you select have the above information –
if any is missing for a particular house, choose another house. If year built is given, subtract that from 2022 to
get age in years.
Example:
address
1060 Blackhawk Dr, University Park, IL 60484
area
(sq.ft)
1,870
number of
bedrooms
age
asking
(years)
price
3
20
135,000
2. Run a simple regression analysis on ‘asking price’ versus ‘area in square feet’ (‘area in square feet’ being the
independent variable). You may use Excel or StatCrunch. Create a report with
page 1: the data table (address, area in square feet, number of bedrooms, age, asking price)
page 2: scatter plot of ‘asking price’ versus ‘area in square feet’ and
page 3: a narrative with
the line of best fit (the equation),
a comment about how good the fit is (like: poor, fair, good),
a numerical measure for the fit,
an interpretation of the above measure of “fit”,
whether the slope is significant,
an interpretation of the slope,
a 95% confidence interval for the slope,
and an estimate of the ‘asking price’ for a house with an area of 2100 square feet.
(5% Extra credit: using StatCrunch, find the 95% confidence interval for the mean price of houses with
an area of 2100 square feet, and a prediction interval for the price of a house with 2100 square feet).
3. Run a multiple regression analysis on ‘asking price’ versus ‘area in square feet’, ‘number of bedrooms’, and
‘age’. Add one more page to the previous report:
page 4: a narrative with
the line of best fit,
a comment about how good the fit is (like: poor, fair, good),
a numerical measure for the fit,
an interpretation of the above measure of “fit”,
whether at least one slope is (significantly) different from zero,
and if so, whether each of the individual slopes is different from zero,
a 95% confidence interval for each slope,
and an estimate of the ‘asking price’ for a house with an area of 2100 square feet, 4 bedrooms,
and 20 years of age.
(5% Extra credit: using StatCrunch, find the 95% confidence interval for the mean price of houses with
an area of 2100 square feet, 4 bedrooms, and 20 years of age; and a prediction interval of the price of
a house with an area of 2100 square feet, 4 bedrooms, and 20 years of age; and check for
multicollinearity among the three independent variables.)
Submit the report (a Word file – not pdf (so that the instructor can copy the table for analysis), with tables and
charts copied and pasted into it) with a cover page which has your name.
