Overview (MRSETS command)

The MRSETS command defines and manages multiple response sets. The set definitions are saved in the IBM® SPSS® Statistics data file, so they are available whenever the file is in use. Multiple response sets can be used in the GGRAPH and CTABLES (Tables option) commands. Two types of multiple response sets can be defined:

• Multiple dichotomy (MD) groups combine variables so that each variable becomes a category in the group. For example, take five variables that ask for yes/no responses to the questions: Do you get news from the Internet? Do you get news from the radio? Do you get news from television? Do you get news from news magazines? Do you get news from newspapers? These variables are coded 1 for yes and 0 for no. A multiple dichotomy group combines the five variables into a single variable with five categories in which a respondent could be counted zero to five times, depending on how many of the five elementary variables contain a 1 for that respondent. It is not required that the elementary variables be dichotomous. If the five elementary variables had the values 1 for regularly, 2 for occasionally, and 3 for never, it would still be possible to create a multiple dichotomy group that counts the variables with 1’s and ignores the other responses.
• Multiple category (MC) groups combine variables that have identical categories. For example, suppose that instead of having five yes/no questions for the five news sources, there are three variables, each coded 1 = Internet, 2 = radio, 3 = television, 4 = magazines, and 5 = newspapers. For each variable, a respondent could select one of these values. In a multiple category group based on these variables, a respondent could be counted zero to three times, once for each variable for which he or she selected a news source. For this sort of multiple response group, it is important that all of the source variables have the same set of values and value labels and the same missing values.

The MRSETS command also allows you to delete sets and to display information about the sets in the data file.