A dummy is a nominal variable with the values 0 and 1. It can be seen as an indicator variable: applicable or not applicable.
The difference between a dichotomy and a dummy
The use of dummy variables
A dummy can be seen as an indicator variable. A 1 means applicable and 0 means not applicable. For instance a question like Did you pass the test? yes / no, is a dummy variable. A question like Do you live in ? yes / no, is a dummy variable. Likewise gender can be code: Are you male? yes / no. It’s a bit strange to say gender is a dummy variable, but you might get the idea. For me it is equal to state the question Are you female? yes / no. This is a dummy variable as well.
Tickle the objects you have at home:[ ] a television
[ ] a radio
[ ] internet
[ ] a washing machine
[ ] a boiler
This isn’t one question, but in reality there are five. And all questions are a dichotomy. Even better, it delivers five dummy-variables. So you need fives columns in your database and the answers should be coded as 1 (= tickled) and 0 (= not tickled). Most programmes for online survey’s code not tickled as blank. These blanks should be replaced with zeroes because a blank means missing and that’s not the same as not tickled. I hope providers for surveys are willing to repair this error in their programme.
A dummy-variable can be used in t-tests, Mann-Whitney test and in regression. Because nominal variables that are not a dichotomy cannot be used in regression, very often such variables are transformed into a list of dummy variables. For instance the multiple choice question:
In which country do you live?0 Germany
0 Elsewhere, .....................................
can be replaced by:
In which country do you live?[ ] Germany
[ ] French
[ ] Spain
[ ] Elsewhere
Now you have created 4 dummy variables. Be aware not to use all four dummies in the analysis at the same time. If a person is not living in Germany, French or Spain it is obvious he lives elsewhere. So three dummy-variables are enough to take into account.