Quick Tab

Select the Quick tab of the Link analysis Startup Panel to access options to select variables and their types for the analysis. You can also exclude certain data cases of your choice using the Select option located on the Startup Panel.

Non-sequential association analysis

Select this check box to perform a nonsequential association analysis.

Variables

Click this button to display a variable selection dialog where you can select variables for the analysis. For sequence analysis (when the Non-sequential association analysis check box is not selected) a five variable selection dialog will be displayed where you can specify the following:

Sequence ID variable

Variables selected in this list will be used as sequence IDs. IDs can be numeric integers or text. Each unique value of the ID variable indicates a unit. For example, in a web log analysis each ID can be uniquely assigned to a user (who is identified by a unique IP address).

Time variable (optional)

Use this list to specify a variable as time (measured in arbitrary units), which is used to arrange transactions in a sequential order. The time variable can have arbitrary units and negative values. A transaction occurring at, for example, time –10, anticipates a transaction occurring at time 10 by 20 units. These units can be seconds, hours, days, etc.

Multiple response vars

Use this list to specify the multiple response variables. See Multiple response variables for further details.

Multiple dichotomy vars

Use this list to specify the multiple dichotomy variables. See Multiple dichotomies for further details.

Continuous vars

Use this list to specify the continuous variables for the analysis. Continuous variables involve quantities that can assume values that can be arranged or sorted in a natural numeric order, i.e., 0.1, 0.2, 0.3… etc. STATISTICA SAL transforms continuous variables into discrete variables by dividing them into a number of segments so the sequence/association analysis can be applied (10 by default).

For association analyses (when the Non-sequential association analysis check box is selected), a four variable selection dialog will be displayed where you can select Multiple response, Multiple dichotomy, Continuous (see above), and Transaction ID variables:

Transaction ID variable

Use this list to select a single variable as the transaction ID variable. This variable will be used to label/identify transactions in the data set. IDs can be numeric integers or text. Each unique value of the ID variable indicates a unit. For example, in a market basket scenario, each ID should represent a single customer (who might be uniquely identified by a credit card number or even a name).

Select single code for all dich. vars.

Select this check box to specify a single code for the dichotomy variables. This option is not available when no dichotomy variables are selected (see the Variables option above). When the check box is selected, a standard code selection dialog can be displayed (by double-clicking on the box adjacent to the check box), where you can select a single code common to all dichotomy variables. Codes must be integer values or text labels (and can be dates, times, etc.).

Multiple dichotomy codes

This button is available when at least one dichotomy variable is included in the analysis (see the Variables option described above) and when the Select single code for all dich. vars (see above) check box is not selected. Click this button to display a standard STATISTICA Multiple Code selection dialog where you select individual codes for the dichotomy variables. Codes must be integer values or text labels (and can be dates, times, etc.).