2. GM 533 Course Project Tips
• Note that I am using a different data set
• These are just ideas
• No guarantee that “this” is what your
instructor is looking for
• Be creative
• Be innovative
3. GM 533 Course Project Tips
• Data Set
– 88 Home Sales
• Square footage
• # Bedrooms
• # Bathrooms
• Type of heating system
• Style of home (from 3 different)
• Garage capacity (# cars)
• Age of home (years)
• Fireplace (yes or no)
• Basement (yes or no)
• Price (in thousands)
• School District (from 2 different)
4. GM 533 Course Project Tips
• Approach
– I will analyze the data first all together
– By School district
– By School district (and type of home)
• Specifics
– Collect descriptive statistics on the set and subsets
– Perform multiple regression analysis on the set and
subsets
• Goal
– Identification of a usable cost model for the
set/subsets, deciding which is most applicable to best
predict the price of a comparable home.
5. GM 533 Course Project Tips
• First Data Set (all)
– 88 homes
• Square footage (Mean = 1735, Std. Dev. = 443.2)
• Bedrooms (Mean = 3.6. Std. Dev. =0.72)
• Bathrooms (Mean = 2.6, Std. Dev. = 0.56)
• Age (Mean = 11.2 years, Std. Dev. = 4.6)
• Price (Mean = $96.8K, Std. Dev. = $26.0K)
• 39 in Eastville School District
– 22 trilevel, 10 two-story, 7 ranch
• 49 in Apple Valley School District
– 11 trilevel, 10 two-story, 28 ranch
6. GM 533 Course Project Tips
• Multiple regression equation for first data set (all)
• PRICE = -12.4694 + 0.0338868 SQ_FT + 4.71321 BEDS
+ 1.47474 BATHS + 14.4475 GARAGE + 13.2394
BASEMENT - 1.37587 AGE + 5.17859 FIRE
– Note that Style, Heat and School variables were treated as
categorical predictors using the General Regression
function in Minitab
• Summary of Model
S = 12.1751 R-Sq = 79.87% R-Sq(adj) = 78.11%
PRESS = 15076.0 R-Sq(pred) = 74.41%
7. GM 533 Course Project Tips
• Second Data Set (Eastville School District)
– 39 homes
• Square footage (Mean = 1627.6, Std. Dev. = 416.7)
• Bedrooms (Mean = 3.5. Std. Dev. =0.68)
• Bathrooms (Mean = 2.5, Std. Dev. = 0.51)
• Age (Mean = 12.3 years, Std. Dev. = 4.3)
• Price (Mean = $85.4K, Std. Dev. = $18.8K)
• Style of Homes
– 22 trilevel, 10 two-story, 7 ranch
8. GM 533 Course Project Tips
• Multiple regression equation for second data set
(Eastville only)
• PRICE = 26.2142 + 0.0208405 SQ_FT + 1.57542 BEDS +
5.60204 BATHS + 5.81769 GARAGE + 10.3948
BASEMENT - 1.68662 AGE + 8.97767 FIRE
– Note that Style and Heat variables were treated as
categorical predictors using the General Regression in
Minitab
• Summary of Model
S = 8.91921 R-Sq = 81.60% R-Sq(adj) = 77.45%
PRESS = 3987.37 R-Sq(pred) = 70.25%
9. GM 533 Course Project Tips
• Second Data Set (Apple Valley School District)
– 49 homes
• Square footage (Mean = 1821.1, Std. Dev. = 449)
• Bedrooms (Mean = 3.7. Std. Dev. =0.75)
• Bathrooms (Mean = 2.7, Std. Dev. = 0.57)
• Age (Mean = 10.3 years, Std. Dev. = 4.7)
• Price (Mean = $106.0K, Std. Dev. = $27.5K)
• Style of homes
– 11 trilevel, 10 two-story, 28 ranch
10. GM 533 Course Project Tips
• Multiple regression equation for second data set
(Apple Valley only)
• PRICE = -35.6487 + 0.0445278 SQ_FT + 6.47639 BEDS
- 3.62598 BATHS + 16.59 GARAGE + 28.9237
BASEMENT - 0.881061 AGE - 6.61155 FIRE
– Note that Style and Heat variables were treated as
categorical predictors using the General Regression in
Minitab
• Summary of Model
S = 11.3604 R-Sq = 85.41% R-Sq(adj) = 82.92%
PRESS = 9248.79 R-Sq(pred) = 74.50%
11. GM 533 Course Project Tips
• Third Data Set (Eastville School
District, trilevel)
– 22 homes
• Square footage (Mean = 1650.6, Std. Dev. = 305.0)
• Bedrooms (Mean = 3.5. Std. Dev. =0.67)
• Bathrooms (Mean = 2.4, Std. Dev. = 0.50)
• Age (Mean = 12.0 years, Std. Dev. = 4.3)
• Price (Mean = $80.2K, Std. Dev. = $13.5K)
12. GM 533 Course Project Tips
• Multiple regression equation for second data set
(Eastville trilevel only)
• PRICE = 24.3398 + 0.0147668 SQ_FT + 3.44476
BEDS + 6.08874 BATHS + 6.44227 GARAGE +
4.07226 BASEMENT - 1.40108 AGE + 8.85074
FIRE
– Note that Heat variable was treated as categorical
predictors using the General Regression in Minitab
• Summary of Model
S = 8.77049 R-Sq = 71.76% R-Sq(adj) =
57.65% PRESS = 2919.42 R-Sq(pred) = 23.45%
13. GM 533 Course Project Tips
• Fourth Data Set (Eastville School District, two
story)
– 10 homes
• Square footage (Mean = 1257, Std. Dev. = 399.0)
• Bedrooms (Mean = 3.7. Std. Dev. =0.82)
• Bathrooms (Mean = 2.2, Std. Dev. = 0.42)
• Age (Mean = 15.0 years, Std. Dev. = 3.8)
• Price (Mean = $78.8K, Std. Dev. = $16.9K)
14. GM 533 Course Project Tips
• Multiple regression equation for second data set
(Eastville two story only)
• PRICE = 44.8562 + 0.0327335 SQ_FT - 1.53464 BEDS +
6.71405 BATHS - 1.08326 AGE - 0.0602882 FIRE
– Note that Heat variable was treated as categorical
predictors using the General Regression in Minitab
• Summary of Model
S = 4.72808 R-Sq = 96.53% R-Sq(adj) = 92.19%
PRESS = 595.740 R-Sq(pred) = 76.87%
16. GM 533 Course Project Tips
• Multiple regression equation for second data set
(Eastville ranch only)
• PRICE = 171.831 - 0.0357418 SQ_FT + 18.6344 BEDS -
5.53515 AGE
– Note that Heat variable was treated as categorical
predictors using the General Regression in Minitab, also
that bathrooms and fireplace variable dropped out since
they were all the same
• Summary of Model
S = 3.08970 R-Sq = 98.02% R-Sq(adj) = 96.05%
PRESS = 158.898 R-Sq(pred) = 89.03%
17. GM 533 Course Project Tips
• Continue this process for all three styles of homes in
the Apple Valley School District
• Decide which model or models are the best approach
for YOU
– You are the decision maker
• In your report, present the models that represent the
home prices best, the variables that matter, etc.
• Leave the majority of the analysis work in the
appendices
• Put complete data set in your Appendices (probably
first)
18. GM 533 Course Project Tips
• Decide how you will break your data down
• Describe your data as you separated it (What
does a typical home look like – use descriptive
statistics, etc.)
• Present model
• Use your other business knowledge to present
innovative ideas as to what other factors may
play a role and that should be considered
19. GM 533 Course Project Tips
• Include a few graphs in your report
– Bar Graphs?
– Pie Charts?
– Think out of the box
• This is a business report
20. GM 533 Course Project Tips
• I will be back on Monday evening at the same
time (in the Week 6 area) to present part of a
final project that “I think” would work
– Again please note these are my ideas and that I do
NOT grade your final project
– See you Monday evening.