Regression Modeling

by


Assignment Content

  1. Purpose 
    This assignment provides an opportunity to develop, evaluate, and apply bivariate and multivariate linear regression models.

    Resources: Microsoft Excel®, DAT565_v3_Wk5_Data_File

    Instructions:
    The Excel file for this assignment contains a database with information about the tax assessment value assigned to medical office buildings in a city. The following is a list of the variables in the database:

    • FloorArea: square feet of floor space
    • Offices: number of offices in the building
    • Entrances: number of customer entrances
    • Age: age of the building (years)
    • AssessedValue: tax assessment value (thousands of dollars)
    • Use the data to construct a model that predicts the tax assessment value assigned to medical office buildings with specific characteristics.

    • Construct a scatter plot in Excel with FloorArea as the independent variable and AssessmentValue as the dependent variable. Insert the bivariate linear regression equation and r^2 in your graph. Do you observe a linear relationship between the 2 variables?
    • Use Excel’s Analysis ToolPak to conduct a regression analysis of FloorArea and AssessmentValue. Is FloorArea a significant predictor of AssessmentValue?
    • Construct a scatter plot in Excel with Age as the independent variable and AssessmentValue as the dependent variable. Insert the bivariate linear regression equation and r^2 in your graph. Do you observe a linear relationship between the 2 variables?
    • Use Excel’s Analysis ToolPak to conduct a regression analysis of Age and Assessment Value. Is Age a significant predictor of AssessmentValue?
    • Construct a multiple regression model.
    • Use Excel’s Analysis ToolPak to conduct a regression analysis with AssessmentValue as the dependent variable and FloorArea, Offices, Entrances, and Age as independent variables. What is the overall fit r^2? What is the adjusted r^2?
    • Which predictors are considered significant if we work with α=0.05? Which predictors can be eliminated?
    • What is the final model if we only use FloorArea and Offices as predictors?
    • Suppose our final model is:
    • AssessedValue = 115.9 + 0.26 x FloorArea + 78.34 x Offices
    • What wouldbe the assessed value of a medical office building with a floor area of 3500 sq. ft., 2 offices, that was built 15 years ago? Is this assessed value consistent with what appears in the database?