A dichotomous variable is a variable that has exactly two possible values. These values represent mutually exclusive categories, such as yes/no, present/absent, or success/failure. Dichotomous variables are also called binary variables.
They often appear in medical, social, and behavioral research where the outcome is a two-way classification. In data analysis, dichotomous variables are often coded numerically (e.g., 0 and 1) to simplify modeling in procedures like logistic regression or hypothesis testing.
Here are some common examples:
| Variable Name | Possible Values |
| Gender Identity | Male, Female |
| Test Result | Positive, Negative |
| Voted in Election | Yes, No |
| Device Working | Working, Not Working |
| Smoker Status | Smoker, Non-Smoker |
In some cases, researchers recode a continuous variable into a dichotomous one to simplify analysis. For example, analysts can convert income into “above median” and “below median.” While this can make results easier to interpret, it also reduces precision and may obscure important patterns in the data.
For categorical variables with more than two categories, analysts often create multiple dichotomous variables called indicator variables (also known as dummy variables). This allows each category to be represented in a regression model using a series of 0/1 variables. It’s a standard method for including categorical information in linear or logistic regression.
« Back to Glossary Index