NAME: Modeling home prices using realtor data
TYPE: Random sample
SIZE: 76 observations, 19 variables

DESCRIPTIVE ABSTRACT:
The data file contains information on 76 single-family homes in
Eugene, Oregon during 2005.  This dataset is suitable for a complete
multiple linear regression analysis of home price data that covers
many of the usual regression topics, including interaction and
predictor transformations.  Whereas realtors use experience and local
knowledge to subjectively value a house based on its characteristics
(size, amenities, location, etc.) and the prices of similar houses
nearby, regression analysis can provide an alternative that more
objectively models local house prices using these same data.

SOURCES:
The data were provided by Victoria Whitman, a realtor in Eugene, in
2005.  The data were used in a case study in Pardoe (2006).

VARIABLE DESCRIPTIONS:
id = ID number
Price = sale price (thousands of dollars)
Size = floor size (thousands of square feet) 
Lot = lot size category (from 1 to 11)
Bath = number of bathrooms (with half-bathrooms counting as 0.1)
Bed = number of bedrooms (between 2 and 6)
BathBed = interaction of Bath times Bed
Year = year built
Age = age (standardized: (Year-1970)/10)
Agesq = Age squared
Garage = garage size (0, 1, 2, or 3 cars)
Status = act (active listing), pen (pending sale), or sld (sold)
Active = indicator for active listing (reference: pending or sold)
Elem = nearest elementary school (edgewood, edison, harris, adams,
crest, or parker)
Edison = indicator for Edison Elementary (reference: Edgewood
Elementary)
Harris = indicator for Harris Elementary (reference: Edgewood
Elementary)
Adams = indicator for Adams Elementary (reference: Edgewood
Elementary)
Crest = indicator for Crest Elementary (reference: Edgewood
Elementary)
Parker = indicator for Parker Elementary (reference: Edgewood
Elementary)

SPECIAL NOTES:
None.

STORY BEHIND THE DATA:
The data file contains information on 76 single-family homes in
Eugene, Oregon during 2005.  At the time the data were collected, the
data submitter was preparing to place his house on the market and it
was important to come up with a reasonable asking price.  Whereas
realtors use experience and local knowledge to subjectively value a
house based on its characteristics (size, amenities, location, etc.)
and the prices of similar houses nearby, regression analysis provides
an alternative that more objectively models local house prices using
these same data.  Better still, realtor experience can help guide the
modeling process to fine-tune a final predictive model.  For example,
both realtor experience and regression modeling results suggest the
need for a BathBed interaction term and an Age-squared transformation
in the model.

PEDAGOGICAL NOTES:
It can be challenging when teaching regression concepts to find
interesting real-life datasets that allow analyses that put all the
concepts together in one large example.  For example, concepts like
interaction and predictor transformations are often illustrated
through small-scale, unrealistic examples with just one or two
predictor variables that make it difficult for students to appreciate
how these concepts might be applied in more realistic multi-variable
problems.  This dataset addresses this challenge by allowing for a
complete multiple linear regression analysis of home price data that
covers many of the usual regression topics, including interaction and
predictor transformations.  The statistical ideas discussed range from
those suitable for a second college statistics course to those
typically found in more advanced linear regression courses.

REFERENCES:
Pardoe, I. (2006). Applied Regression Modeling: A Business Approach.
Hoboken, NJ: Wiley.

SUBMITTED BY:
Iain Pardoe
University of Oregon
Lundquist College of Business, 1208 University of Oregon, Eugene, OR
97403, USA.
ipardoe@lcbmail.uoregon.edu