The objective of this project is to develop and estimate multiple regression models of housing values for properties affected by the widening of Springbank Drive. The City of London plans to compensate homeowners for the lost frontage area and the increased traffic resulting from the road widening. By utilizing multiple linear regression models, we aim to objectively determine the appropriate compensation for the affected homeowners, including the $12,000 lump sum.
To generate the regression models, a sample of 104 homes located around Springbank Drive, sold between January 1998 and May 2003, was used. The following descriptive statistics were derived from this sample:
Table 1: Descriptive Statistics of 104 Homes around Springbank Drive
It is important to note that the scale of measurements varies across variables, making direct comparisons between them less meaningful. However, binary variables and dummy variables provide valuable insights, representing qualitative measures such as the view or condition of the home. These variables play a significant role in home valuation and should be included in the models.
Our methodology focuses on multiple regression as it aligns with the specific context of the widening of Springbank Drive. Variables such as traffic count and the presence of a four-lane road directly relate to the compensation resulting from the road widening. Multiple regression allows for a more accurate valuation by considering each variable's impact while holding others constant.
The chosen variables represent factors that most significantly influence home values in the Springbank Drive area. These factors include the home's size, condition, amenities, and relevant property characteristics. Additionally, the assumption of normality in the data enables the soundness of using multiple regression for determining compensation figures.
The coefficients obtained from the regression models provide valuable insights into the valuation of homes affected by the widening of Springbank Drive. Key findings include:
Based on the regression models, two formulas can be considered for determining the final compensation figures for the impacted homes:
Model 1: Traffic Count Model
Compensation = $12,000 (Lump Sum) + $38.55 × (Sq. Ft. of Frontage Lost) + $7,967 (If home now on Four Lane Road)
Model 2: Four Lane Road Model
Compensation = $12,000 (Lump Sum) + $40.68 × (Lot Frontage Lost) + $0.84 × (33,000 - old traffic count)
These formulas provide a fair and objective approach to calculating the appropriate compensation for the homeowners based on the specific variables and their associated coefficients.
One limitation of the project is the sample size of 104 observations relative to the number of potential variables that could affect home prices. Due to this constraint, only a limited number of variables could be included in the regression models. Furthermore, the adjusted R-squared value indicates that additional variables beyond the included ones may contribute to explaining.