The way INTERACT interprets the available information, during the statistical calculations for Frequency of occurrence and the Durations diagram, can be adjusted to your needs:
▪Click General - Program - Settings 
to open the 'Settings' dialog.
▪Switch to the tab Calculation and change the settings according to your needs.
Calculation settings
These settings apply to 'duration' and 'frequency' pie-charts available in the Visualization group:
oDo not count empty codings: Only Codes actually logged within the selected Class, are counted and included in the statistical calculations (see next option).
oCount all contents: Empty cells (Events with no Code for a particular Class) are counted as well. Therefore the 100% base includes all Events of this Class within the selected DataSet(s). The resulting graph contains a value for "not coded".
oCount "coded" vs. "not coded" only: The resulting graph contains only two values. One for the number of Codes given (independent of the kind of Code logged, within the selected Class and DataSet(s)) and one for the Events without a Code for the selected Class and DataSet(s).
DataSet time boundaries
Another important setting concerns the way INTERACT handles the time borders of each DataSet. Events outside the time borders of the DataSet are not displayed in the Timeline Chart and thus will not be considered in the Time Line-based statistics.
oAsk before correction: If Events are found, that do not fit within the DataSet borders a dialog appears to ask you for permission to correct the borders.
▪Read Using DataSet Time Borders for details.
oCorrect without asking: Default setting; whenever the Timeline Chart is opened INTERACT checks the time boundaries per DataSet and corrects them automatically, if Events outside the existing borders are found.
Statistical Results
DataSet-based statistical results from routines per default exclude DataSets that do not hold the required Code. If you want all DataSets to appear in the results dialog:
▪Select Show DataSets without statistical results in reports.