Editing Multiple Response Sets

You can add or edit multiple response sets from any node that includes a Filter tab by clicking the Filter button menu in the upper left corner and choosing Edit Multiple Response Sets.

Multiple response sets are used to record data that can have more than one value for each case—for example, when asking survey respondents which museums they have visited or which magazines they read. Multiple response sets can be imported into IBM® SPSS® Modeler using a Data Collection source node or a Statistics File source node and can be defined in IBM SPSS Modeler using a Filter node.

Click New to create a new multiple response set, or click Edit to modify an existing set.

Name and label. Specifies the name and description for the set.

Type. Multiple response questions can be handled in one of two ways:

• Multiple dichotomy set. A separate flag field is created for each possible response; thus, if there are 10 magazines, there are 10 flag fields, each of which can have values, such as 0 or 1 for true or false. The counted value allows you to specify which value is counted as true. This method is useful when you want to allow respondents to choose all options that apply.
• Multiple category set. A nominal field is created for each response up to the maximum number of answers from a given respondent. Each nominal field has values representing the possible answers, such as 1 for Time, 2 for Newsweek, and 3 for PC Week. This method is most useful when you want to limit the number of answers—for example, when asking respondents to choose the three magazines they read most frequently.

Fields in set. Use the icons on the right to add or remove fields.

Comments

• All fields included in a multiple response set must have the same storage.
• Sets are distinct from the fields they include. For example, deleting a set will not cause the fields it includes to be deleted—merely the links between those fields. The set is still visible upstream from the point of deletion but is not visible downstream.
• If fields are renamed using a Filter node (directly on the tab or by choosing the Rename for IBM SPSS Statistics, Truncate, or Anonymize options on the Filter menu), any references to these fields used in multiple response sets will also be updated. However, any fields in a multiple response set that are dropped by the Filter node will not be removed from the multiple response set. Such fields, although no longer visible in the stream, are still referenced by the multiple response set; this could be a consideration when exporting, for example.