8210 11 assgn

SkillBuilder18_InterpretingRegressionCoefficientsforDummy-CodedVariables-HowtoCreateDummy-CodedVariables.html

Home >Education homework help >8210 11 assgn

Magnifying glassSkill Builder 18: Interpreting Regression Coefficients for Dummy-Coded VariablesSKIP TO TOPIC60% COMPLETE 60% COMPLETE Magnifying glassSkill Builder 18: Interpreting Regression Coefficients for Dummy-Coded VariablesSKIP TO TOPIC60% COMPLETE 60% COMPLETE 60% COMPLETE 60% COMPLETE

Three vertical lines aligned to the leftInterpreting Regression Models with Dummy-Coded VariablesInterpreting Regression Models with Dummy-Coded Variables 100 Percent Complete A circle with a colored border representing one's progress through a lesson.

Three vertical lines aligned to the leftHow to Create Dummy-Coded VariablesHow to Create Dummy-Coded Variables 100 Percent Complete A circle with a colored border representing one's progress through a lesson.

Three vertical lines aligned to the leftInterpreting the Coefficients for Dummy-Coded VariablesInterpreting the Coefficients for Dummy-Coded Variables 100 Percent Complete A circle with a colored border representing one's progress through a lesson.

Three vertical lines aligned to the leftModule Summary and QuizModule Summary and Quiz 5 Percent Complete A circle with a colored border representing one's progress through a lesson.

Three vertical lines aligned to the leftGlossaryGlossary 0 Percent Complete A circle with a colored border representing one's progress through a lesson. EXIT SKILL BUILDER Topic 1 - Interpreting Regression Models with Dummy-Coded Variables EXIT SKILL BUILDER

How to Create Dummy-Coded Variables

by Robin KouvarasRobin Kouvaras Topic 2 of 5

Learning Objective:Interpret regression models with dummy-coded variables.

Learning Objective: Interpret regression models with dummy-coded variables.

How to Create Dummy-Coded Variables

Dummy-coded variables are created by only using the values of 0 and 1. The general rule used for dummy coding is that you need one (1) fewer dummy-coded variables than you have groups (# total groups – 1). So, for our variable of marital status, we would need two (2) dummy-coded variables because we have chosen to focus on three (3) marital status groups (3 – 2 = 1). The group for which we do not create a dummy-coded variable is typically called the reference category. Often the reference category will be the one that researchers want to compare to other groups. For our research, we might choose “married” as our reference category if we want to compare non-married individuals to married individuals.

Before we conduct our regression analyses in SPSS, then, we will need to create two (2) dummy-coded variables for marital status:

one variable for the divorced group

one variable for the never-married group

We will use a 1 to indicate membership to that category (e.g., to indicate that someone is divorced for the “divorced” dummy-coded variable) and 0 to indicate non-membership.

The table below shows how we would dummy-code our marital status variables.

Table with three columns. The first column is marital status, the second column is divorced, and the third column is never married. The first row is 1 (married), 0, 0. The second row is 2 (divorced), 1, 0. The third row is 3 (never married), 0, 1." title="Table with three columns. The first column is marital status, the second column is divorced, and the third column is never married. The first row is 1 (married), 0, 0. The second row is 2 (divorced), 1, 0. The third row is 3 (never married), 0, 1.

Notice the Following

If the original value for an individual’s marital status is a 1 (indicating married), that individual would have a 0 for the “divorced” variable and a 0 for the “never married” variable. This is because they are not a “member” of either of these groups, they are not divorced, and they are not in the never-married category. This same logic holds for the remaining two (2) values of marital status. If an individual is divorced, they get a 1 for the divorced group, for example, and a 0 for the never-married group.

Also, note that each individual in the data set will have a value (either a 0 or a 1) for each dummy-coded variable that the researcher creates.

Suppose the researcher decides to add an additional marital status group (separated), so that she now has the following marital status groups: married, divorced, never married, and separated.

Hint: Count the number of groups you have and subtract 1.

How many dummy-coded variables would the researcher need to create for her regression model?

5

2

3

4
SUBMIT Incorrect TAKE AGAIN Topic 3 - Interpreting the Coefficients for Dummy-Coded Variables Caret pointing down