Observing multiple subjects is a common situation in behavioral research. Still, it has a sheer endless number of possible implementations.
Note: DO NOT try to squeeze different types of information into a lexical chain, just because it seems possible. Especially if part of the information changes often without affecting the rest. Coding these different types of information fully independent at first is often your better choice. The Pattern Segmentation routine can turn all these independent observations into Event that look like they have been coded using a lexical chain.
What setup works best for you depends on a range of variable characteristics. We have prepared the following basic setups in detailed examples:
 | Observer One Subject per Pass - As the title suggests, we presume you observe each subject seperately. For this setup it is technically irrelevant how many subjects there are and whether you observe all different types of behavior in one pass or in multiple passes. |
| Sequential Observations for Multiple Subjects - In this example we presume something like a conversation where the participants are supposed to act sequentially, so some one says something about a specific topic and to another subject or the group as a whole and after that, another subject does something similar. |
| Simultaneous Observation of Multiple Subjects - If you want to observe the activities of multiple subjects in one pass, you probably concentrate on a single type of behavior. Other, even more complex situations can be implemented, but these are not suitable for general examples. |
Note for all methods: You will probably need different file structures to find all the answers. For this reason, INTERACT offers a range of restructuring commands.
Therefore, you might need to save different versions of your data for evaluation, because certain questions are easier to answer from a different structure.