Regression Models with Interaction Between Variables - PowerPoint PPT Presentation

1 / 15
About This Presentation
Title:

Regression Models with Interaction Between Variables

Description:

When Yrs Exper = 4 & Gender Male (0) YrsExper*Female = 0. Regression Model with Interaction. y = 1.5278x 30.43. Male. y = 0.28x 34.528. Female ... – PowerPoint PPT presentation

Number of Views:507
Avg rating:3.0/5.0
Slides: 16
Provided by: hrcla
Category:

less

Transcript and Presenter's Notes

Title: Regression Models with Interaction Between Variables


1
Regression Models with Interaction Between
Variables
Topics Motivational Example Modeling
Interaction Effect Interpreting Slope Coefficient
for Interaction Variables Interaction with two
quantitative Vars. Implementing Interaction
Variables in StatTools
2
Problem Scenario
  • The Fifth National Bank is facing a
    gender-discrimination suit. The charge is that
    its female employees receive substantially
    smaller salaries than its male employees.

3
Investigating Differences Due to Gender
  • It is customary for salaries to be pegged to
    years of experience
  • Is there any difference between females and males
    in salary increases with years of experience?

4
Salary Variation with Years of Experience Males
vs. Females
5
Salary Variation with Gender 4 vs. 8 Years
Experience
6
Interaction Effect
  • From the graphs we conclude
  • gender moderates the effect of years of
    experience on salary
  • gender has some effect on salary and this is
    moderated by years of experience
  • We say that gender interacts with years of
    experience

7
Modeling Interaction Effect
  • Introduce an extra variable into the model as
    product of interacting variables
  • YrsExperFemale
  • When Yrs Exper 4 Gender Male (0)
  • YrsExperFemale 0

8
Regression Model with Interaction
y 0.28x 34.528 Female
y 1.5278x 30.43 Male
9
Interpretation of Interaction Regression Slope
coefficient
  • Difference in slopes for Salary vs. YrsEper.
    Simple regression Male Female 1.5278 0.28
    1.248
  • Coefficient for YrsExperfemale (-1.248) is the
    salary reduction per extra year of experience for
    females compared to males

10
Interaction between Quantitative Variables
  • A restaurant chain wants to investigate the
    relationship between shop revenue (000s) and the
    median household income (000s) mean age of
    children in the neighborhood
  • Is there interaction between mean age and
    household income?

11
Interaction between Quantitative Variables
12
Interaction between Quantitative Variables
13
Interpretation of Interaction Slope Coefficient
  • For each additional year in mean age of
    neighborhood children, the rate of increase of
    gross revenue per 1,000 increase in median
    household income decreases by 2,151 on average

14
Creating Interaction Variables in StatTools
  • Name the data set in the usual way
  • Place the cursor anywhere in the spreadsheet and
    click on the Data Utilities icon (3rd from left)
  • Select Interaction and click on the down arrow at
    top to select the type of Interaction Between
    (e.g. two numeric or two categorical variables)

15
Creating Interaction Variables in StatTools
  • By clicking, check the box beside the names of
    the variables you want to interact then click
    O.K.
  • Click yes when StatTools warns if you wish to
    continue to insert a new column
  • StatTools will insert the new column with the
    interaction variable next to your data
Write a Comment
User Comments (0)
About PowerShow.com