The best reasons to use Prefixes (works with lexical chained Codes only):
oYou are only interested in statistics per subject
oYou have a lot of identical sub-level Codes that only differ in their Class name
The trick with this Prefix functionality is, that you are able to create context based Class names, for Codes in sub-level Code definition files.
This means, that the Class name of a sub-level Code, depends on the Code selected in the preceding level.
Example: If you want to collect several 'Activity' Codes for, let's say 5 test subjects ('A', 'B', 'C', 'D', 'E'). You want this data (e.g. 'sleeping', 'eating', 'sitting', 'moving') to be entered in a separate column for each Subject: "A_Activity", "B_Activity", "C_Activity", etc.
Without Prefixes, this results in at least 6 linked Code definition files, each of them containing exactly the same Codes, but assigned to another Class. Or you need one large Code definition file, with many, many different Keys for more or less the same Codes.
Note: We presume that you are familiar with the basics of logging Events with INTERACT using either the Coding Mode: Lexical (post-hoc) or the Coding Mode: Complex .
For this example, using Prefixes, we need only 2 Code definition files (even if you have 50 subjects):
▪Create one Code definition file, containing all activities - e.g. Codes with the Class 'Activity':
▪Create another file, containing all subject identifiers; each with a unique entry in the Prefix column (e.g. 'Subject A' with Prefix 'A_', 'Subject B' with Prefix 'B_', etc.):
The Prefix entry ensures that the 'Activity' information is stored in a separate column for each person.
▪Select the option This file contains the "Master Object Codes" at the bottom of the dialog for this top-level file.
▪Link the second-level Activity.ikey file to each of the subjects, using the Lex.Chain column.
During an observational session, this setup results in at least 6 columns called 'Subject', 'A_Activity', 'B_Activity', 'C_Activity' etc.
▪Read the examples within this section because they all use the Prefix attribute.
Some examples are rather easy and straight forward, some are more sophisticated - especially the FACS examples are rather tricky, that is why they are joined in a separate section called Best Practice Examples - Coding Facial Expressions .
Remember: All examples can be adapted to your needs, because the difference lays not so much in WHAT you observe, but rather in HOW you observe!