So far, we have only looked at the data per Subject.
The State-Space-Grid gives us a great view on overlaps between two Classes and multiple Subjects, but if you need more details about interactions between Subjects, like sequences or contingencies, another structure is required.
In order to determine sequences or contingencies between specific behavior of several participants, the data of all subjects ('Speaakers' in this example) must be moved into a single Set per Session.
To do this, you first need to add the not coded, thus missing, 'Subject' code to all Events:
Add Subject-ID to Events
If you collected your data in separate Sets per Subject, like this:
The Events probably do NOT have a Subject code per Event - that would have been a waste during the coding-process.
For some Analysis routines, you need to merge all Events of a single session into the same DataSet, though.
We talk about routines like:
ocreating new Events for co-occurring Codes,
ofinding sequences among subjects, or
oidentifying contingencies (=latency between two Codes),
Before you can do that, you need to add some kind of Subject ID to EACH Event line because otherwise you can no longer tell which Event belongs to whom.
If you did label each Set with a proper Subject-ID, like shown in the picture, adding this information to all the Events within that Set is easy:
▪Select the command Transform - Restructure - Data > Use DataSet description as ID *) and each Event receives a proper Subject-ID instantly:
Only after each Event line received its corresponding 'ParticipantID', it is safe to unite the Events from all participants from the same session.
Note: If you are interested in sequences of behavior across participants, your first might need to Create Code segments before you merge your data. That depends on the structure of collected behavioral data, and the answers you are looking for.
If all ID's are in place and optional segments are created, you can run the Merge DataSets command:
▪Run the command Transform - Restructure - Data 
- Move all Events to a single Dataset (per DataGroup) to merge the Events from all Sets per Group.
The how and why is explained in detail in Session-based Sequence analysis within one of the BEST PRACTICE section.
Other merge possibilities for data within the same file are explained in Unite Events into one DataSet.
If you do not like the Class-Name 'ID', you can rename it to whatever you want easily:
▪Double-Click (in earlier versions you need to tripple-click) the header cell containing 'ID'.
▪Enter the Class name you prefer, for example 'Subject':
▪Hit RETURN to apply the new class name.
Create Code segments
If your data contains overlapping behavior per Subject, and you are interested absolute Code combinations, the routine described in detail in Pattern Segmentation might be just what you need. You should run it before you merge the data from the different Sets!
The command Transform - Events - Create Code segments 
allows you to create a sequential string of Events from just about any number of overlapping Events.
This is extremely helpful if you are interested in Sequence Analysis based on a complex combination of Codes. Such complex situations are hard to describe in a generic way, so if you have trouble getting the results you are interested in, send us data and details and we can give you a helping hand.