Guide to Using Excel 2007 For Basic Statistical Application

Guide to Using Excel 2007 For Basic Statistical Applications

To AccompanyBusiness Statistics: A Decision Making Approach, 8th Ed.Chapter 15:Multiple Regression and Model BuildingByGroebner, Shannon, Fry, & SmithPrentice-Hall Publishing CompanyCopyright, 2011

Chapter 15 Excel Examples

Multiple Regression First City Real EstateMultiple Regression –Dummy Variable First City Real EstateCurvilinear Regression Ashley Investment ServicesInteraction Effects Ashley Investment Services

More Examples

Chapter 15 Excel Examples

Residual Analysis First City Real Estate

Multiple Regression – First City Real Estate

Issue: First City management wishes to build a model that can be used to predict sales prices for residential property. Objective: Use Excel 2007 to build a multiple regression model relating sales price to a set of measurable variables. Data file is FirstCity.xls

Multiple Regression – First City Real Estate

Open the file FirstCity.xls, Sheet Homes-Sample 1

Select the Data Tab

Select Data Analysis

Select Correlation

OK

Multiple Regression – First City Real Estate

Input Range: A1:G320

Grouped by: Columns

New Worksheet Ply: corr-1

OK

Multiple Regression – First City Real Estate

This cell shows the correlation, 0.7477, between Price and Square Feet

Multiple Regression – First City Real Estate

Return to

Homes-Sample 1

Select the Data tab

Select Data Analysis

Select Regression

OK

Multiple Regression – First City Real Estate

Input Y Range: A1:A320

Input X Range: B1:F320

Select labels

New Worksheet Ply: regress-1

OK

Multiple Regression – First City Real Estate

Excel produces the Regression Model

Multiple Regression – First City Real Estate

To determine if multicollinearity is a problem:

Select the Add-Ins tab

Select PHStat

Select Regression

Select Multiple Regression

Multiple Regression – First City Real Estate

Y variable … : A1:A320

X variable … : B1:F320

Select First Cells in both …

Select Regression Statistics …

Select ANOVA and …

Select Variance Inflation …

OK

Multiple Regression – First City Real Estate

PHStat will find Variance Inflation Factors for all independent variables. This slide shows the VIF for Garage # and all other X variables. Click on the sheet tabs X4,X3,X2,X1 to see the VIF for the other variables.

Issue: First City managers wish to improve the model by adding a location variable for the area. Objective: Use Excel 2007 to improve a regression model by adding a dummy variable for the area either foothills or flatland.Data file is First City.xls

Multiple Regression - Dummy Variable -First City Real Estate

Multiple Regression- Dummy Variable – First City Real Estate

Select the Home-Sample 2 sheet

Select the Add-In tab

Select PHStat

Select Regression

Select Multiple Regression

Multiple Regression- Dummy Variable – First City Real Estate

Cut Bedrooms and Bathrooms columns and paste in cell G1. Delete columns C and D so that all data is in connected columns

Select Data tab

Select Regression

Input Y Range: A1 – A320

Input X Range B1-E320

Select Labels

New Worksheet: Regress-2

OK

Multiple Regression- Dummy Variable – First City Real Estate

All variables are significant and have the expected sign

Curvilinear Relationships - Ashley Investment Services

Issue: The director of personnel is trying to determine whether there is a relationship between employee burnout and time spent socializing with co-workers. Objective: Use Excel 2007 to determine whether the relationship between the two measures is statistically significant. Data file is Ashley.xls

Curvilinear Relationships – Ashley Investment Services

Open the file Ashley.xls

Curvilinear Relationships – Ashley Investment Services

Select A1:B21

Select the Insert Tab

Select Scatter Plot

Select the Scatter Plot desired

Curvilinear Relationships – Ashley Investment Services

Curvilinear Relationships – Ashley Investment Services

Select the Data tab

Select Data Analysis

Select Regression

Curvilinear Relationships – Ashley Investment Services

Y Range A1:A21

X Range B1:B21

Labels

New Worksheet Ply: Ashley-1

OK

Curvilinear Relationships – Ashley Investment Services

Curvilinear Relationships – Ashley Investment Services

To develop a nonlinear model, return to the data file. Create a new variable Socialization Squared

Curvilinear Relationships – Ashley Investment Services

Select the Data tab

Select Data Analysis

Select Regression

OK

Curvilinear Relationships – Ashley Investment Services

Input Y Range: A1:A21

Input X Range: B1:C21

Select Labels

New Worksheet Ply: Ashley-2

Curvilinear Relationships – Ashley Investment Services

The output shows the R Square value and the Regression Coefficients.

Interaction Effects - Ashley Investment Services

Issue: The director of personnel is trying to determine whether the model can be improved by separating observations between those taken from men and women. Objective: Use Excel 2007 to determine whether the relationship between the measures can be improved. Data file is Ashley-2.xls

Interaction Effects – Ashley Investment Services

Open file Ashley-2.xls

Insert a new column C as (Socialization Squared) which is Column B squared

Add Column E as Columns B * D

Add Column F as Columns C * D

Interaction Effects – Ashley Investment Services

Using The Insert tab and Chart tools set up a Scatter Plot for one gender

Interaction Effects – Ashley

Investment Services

Add the second gender

Add exponential trend line for male and female

Interaction Effects – Ashley

Investment Services

The regression for the curvilinear model.

Issue: The company is interested in analyzing the residuals of the regression model to determine whether the assumptions of multiple regression are satisfied. Objective: Use Excel 2007 to analyze residuals from a regression model. Data file is First City-3.xls

Residual Analysis - First City Real Estate

Residual Analysis – First City Real Estate

Open the file FirstCity3.xls

Since Excel requires independent variables to be in adjacent columns – cut and paste these columns.

We will be using:

Price

Sq. Feet

Bedrooms

Garage #

Log of Lot Size

Note: I will swap Lot Size and Log Lot Size to simplify the operation

Residual Analysis – First City Real Estate

Select the Data tab

Select Data Analysis

Select Regression

OK

Residual Analysis – First City Real Estate

Define the range for X and Y variables.

Excel gives several options for Residual Analysis but does not have as complete a set as Minitab

Residual Analysis – First City Real Estate

This is the Residual Plot for Square Feet

Residual Analysis – First City Real Estate

While Excel will not automatically generate a histogram of the standardized residuals , one can be created.

Residual Analysis – First City Real Estate

Define Bin values for the Histogram

Residual Analysis – First City Real Estate

Select the Data tab

Select Data Analysis

Select Histogram

On the Histogram Chart identify the data and bins

Select Chart output

Residual Analysis – First City Real Estate