Variable and Case Metadata Overview
Statistica provides you with the capability to append supplementary information called Variable and Case Metadata to your spreadsheets. Variable and Case Metadata is a very simple yet powerful addition to the already extensive Statistica spreadsheet facilities. This Metadata can be used to create a more comprehensive description of your data set, facilitate what-if types of exploratory analyses, and customize the appearance of cases in graphs. A brief description of the new options available as Variable and Case Metadata follow.
Variable Metadata
The modeling or measurement type of a variable can be explicitly defined in order to indicate what analyses and graphs are appropriate for such a variable. Statistica supports the following measurement types: Continuous, Ordinal, and Categorical. These measurement types map directly to subsequent analyses and graphs, identifying appropriate variables in each case (example, variables of type categorical are present within the list of categorical predictors available in a Factorial ANOVA). Note that you can also define a variable as Unspecified, in which case the variable is treated as essentially all measurement types. You can also define a variable as Auto, in which case Statistica automatically determines the measurement type of the variable based on the values of the variable.
Variables specified as Excluded are not available for analyses and graphs. The values of variables marked as Label can be defined as point labels within graphs essentially adding another dimension to your graph.
User defined properties such as notes or documentation for a variable describing the nature of the variable or complex formulae used to compute the variable can also be defined.
Case Metadata/Case States
You can assign case states to cases in order to customize the appearance of points in graphical displays, thus making it very easy to identify influential and interesting points. A wide selection of symbols and colors is available to customize the appearance of selected points.
Not only can case states be assigned in the spreadsheet before a graph is made, they can also be assigned interactively in the graph by the Brushing tool.
The case states assigned in the graph propagate back to the spreadsheet. The ability to assign case states in either the spreadsheet or graph further facilitates the exploratory visual analysis of data.
Additional case states include Hidden and Excluded. The Hidden property allows cases to be used in computations, but not be visible in graphs. Cases marked as Excluded are not used in computations, but are visible in graphs. Marking a case as Hidden and Excluded excludes the case entirely and it does not appear in the graph or its respective computations.