Multiple Response Tables - Quick Tab

Select the Quick tab of the Multiple Response Tables dialog box to access options to specify so-called multiple response variables and multiple dichotomies (see Multiple responses/dichotomies), and compute 1) frequency tables for such variables or 2) compute crosstabulation tables involving such variables (e.g., ordinary categorical variables can be crosstabulated with multiple response variables/dichotomies.)

Specify table (select variables)

Click this button to display the standard variable selection dialog box. Each list will be interpreted as a single multiple response/dichotomy variable or factor. If only a single variable is chosen in a list, then the respective variable will be treated as a standard categorical (grouping) variable in the crosstabulation. By default, 1) short and long names as well as 2) short and long alphanumeric value labels for the factors will be taken from the first variable in each list. Factor names can later be changed (see Long Factor Labels for Multiple Response Tables).

Paired crosstabulation

Select the Paired crosstabulation check box only if there are two or more multiple response variables specified for the current table and if at least two of them consist of identical numbers of actual variables. If so, Statistica will proceed as follows:

• The number of actual variables in each multiple response factor is counted;
• Paired crosstabulation is performed on the first two or more multiple response variables with identical numbers of actual variables.

Thus, for example, if there are three multiple response factors, the first two with 2 actual variables and the third with 3, only the first two factors will be processed as paired factors. The logic of paired crosstabulation of multiple response variables is explained in the overview section (under the header Paired Crosstabulation of Multiple Response Variables).

Name of Factor ( Mult

Resp. Set). By default, the names for factors are taken from the first variable in each respective list. To change the names, simply enter the new name into the edit field next to the respective factor.

No

of Vars. The number of variables that were chosen in each respective list will be displayed here.

Type of Multiple Response Factor

Click these links for an overview of the general logic of multiple response variables and multiple dichotomies.

Multiple dichotomy

Select the Multiple dichotomy option button to treat the respective factor (list of actual variables) as a multiple dichotomy. Be sure that the desired Count value is set (by default 1, see below). All values that do not match the count value will be ignored in the analysis.

Multiple response

Select the Multiple response option button to treat the respective factor (list of actual variables) as a multiple response variable.

Codes. Select the Codes button to display the Select Codes for Multiple Response Sets (Factors) dialog box. Note that this option is only available if you have selected the Multiple response option button for the respective factor. If no Codes are selected, STATISTICA will automatically tabulate all integer values found in the first variable of the respective list. If there is only one variable in the list, the respective variable will be treated (tabulated) as an ordinary categorical (grouping) variable.

Read all text labels. Select this check box to compute the Multiple Response codes for text variables based on their text labels. This resolves the computational discrepancies when two or more variables have instances of the same text labels but different numeric codes.

Keep order as found in data. When this check box is selected, the order of the categorical labels will be in the same sequence as found in the data file; clear it to order them alphabetically (case-insensitive). When there are formatted (e.g., colored or bold) and non-formatted (regular/default) categorical labels with the same text, the instance that appears first in the data file will be used for the categorical labels (for both short and long labels).

Count value

Specify the actual value that will be counted in the multiple response table in the Count value field; that is, STATISTICA will count how frequent this value is present in the different actual variables (levels) that comprise the respective Multiple dichotomy.

Count unique responses only (ignore multiple identical responses)

Select the Count unique responses only (ignore multiple identical responses) check box to ignore the multiple identical responses in the different actual variables comprising a respective multiple response factor. This is the default setting. For example, if a multiple response factor Preferred Sport consists of two actual variables (first preference, second preference) and a respondent listed the same sport for both preferences, then STATISTICA will only count this as one response. (In other popular programs that compute multiple response tables, this respondent would be counted twice; clear this check box if this is what is desired for the current crosstabulation or frequency table.)