A variable with just two categories that reflects only part of the information actually available in a more comprehensive variable. For example, the four-category variable Region (Northeast, Southeast, Central, West) could be the basis for a two-category dummy variable that would distinguish Northeast from all other regions. Dummy variables often come in sets to reflect all of the original information. In our example, the four-category region variable defines four dummy variables: (1) Northeast vs. all other; (2) Southeast vs. all other; (3) Central vs. all other; and (4) West vs. all other. Alternative coding procedures (which are equivalent in terms of explanatory power but which may produce more easily interpretable estimates) are effect coding and orthogonal coefficients.
In regression analysis, a dummy variable (also known as indicator or bound variables) is one that takes the values 0 or 1 to indicate the absence or presence of some categorical effect that may be expected to shift the outcome. For example, in econometric time series analysis, dummy variables may be used to indicate the occurrence of wars, or major strikes.