# Thread: Help with "nominal" variables and dummy coding

1. ## Help with "nominal" variables and dummy coding

Hi.

First of all, I'm sorry for not writing 100% correct as this is not my everyday language

My quetion

I'm working with linear regression in SPSS, and everything has been OK so far, but now I'm asked to implement a nominal variable (Gender) into my model. I know that I need to make ONE new dummy variable, but I'm really not sure about how I should do that?

In my dataset, I have a variable called "Sex" (... along with a few other ratio scaled variables) and it's presently coded as follows:
0 = Male
1 = Female

How should I go forth about creating the dummy variable? Any ideas?
Have in mind that I'm a newbie, so detailed explanations are more then welcome

Many tanks

2. ## Re: Help with "nominal" variables and dummy coding

The dummy variable is almost always (that I've seen) coded 0/1. Since "Sex" is already a 0/1 variable, then you don't really need to re-code it; but if you want to:
Think of the dummy variable as measuring "male-ness", so a guy would score a 1 for male-ness, and a lady would score a 0 for male-ness.
So if Sex=Male, then maleness=1. Else if Sex=Female, then maleness=0.
Usually, people use "Male" instead of maleness for the variable name.

So, we see that if we have a Sex variable with 2 levels, then we make 1 dummy variable. If you ever run into a variable with 5 levels, then you'd need to make 4 dummy variables, etc.

3. ## Re: Help with "nominal" variables and dummy coding

So I dont need two variables? The original "sex" variable AND another dummy variable? It is enough with the one "sex" variable because it's coded 1/0?

4. ## Re: Help with "nominal" variables and dummy coding

Yes, that's correct.

 Tweet

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts