Guided Tour - Working with Keywords
Another common approach to qualitative analysis involves creating a coding scheme and applying those codes to portions of the source data. In Transana, this is accomplished by creating keywords and applying those keywords to clips. (All coding in Transana is done at the clip level.) Once you've done this, you can approach your data in a variety of new ways as you try to make sense of what it has to show you.
The traditional clips described earlier are named and assigned to a collection when they are created, and therefore are situated in theory from the outset. Keywords can be assigned to traditional clips. In other words, traditional clips can be coded.
But once you start creating keywords, a second method of creating clips is available. Quick Clips are clips that are created in association with a keyword rather than a collection, thus allowing for an "open-coding" approach to video and audio data. Quick clips are not named by the researcher during creation and are not initially placed in any theory-bound collection. This allows the researcher to defer these tasks to later in the analytic process, after an initial theoretical understanding has had time to emerge from the data, or to work in an analytic methodology that does not rely on collections to indicate theoretical categorization.
Transana provides a report generator that allows you to see exactly how you've used episodes, collections, clips, and keywords. You can select what data elements, as well as what clips and keywords, should be included in the report, allowing you to craft text-based reports that meet your particular needs.
Transana's text reports can be configured, edited, saved in Rich Text Format, or printed.
Because video and audio files have an underlying time line, a variety of graphical reports about coding are also possible. The keyword map allows you to examine and interpret keyword placement over time.
The keyword maps above allowed Mitchell Nathan to note a persistent pattern in a particular mathematics class. This pair of keyword maps shows that the Initiation - Demonstration - Evaluation/Elaboration pattern Dr. Nathan thought he saw actually occurs 14 times in the course of about 18 minutes of student-led discussion. This map also led to a productive discussion of the distinction between Evaluation and Elaboration, as it was noted that the two almost always co-occur during this class period. Note that in Transana, you can click on any of the bars in the Keyword Map and Transana will load the clip associated with that bar in the map, allowing you to easily examine the source video underlying the data points in the graphical reports.
Nicolas Sheon wanted to help novice peer AIDS educators examine their counseling skills and use of time during counseling sessions. He created a closed coding scheme and trained these peer counselors to code recordings of their sessions using quick clips. The counselors (and their supervisors) then use Transana's series keyword sequence maps to do self-evaluation and to explore changes in their counseling practice over time.
(The graph above is based on artificial data due to the sensitive nature of Dr. Sheon's data and probably over-states the actual results, but I believe it captures the spirit of his work. The Transana development team offers its sincere thanks to Dr. Sheon and the Center for Aids Prevention Studies for funding the development of this functionality within Transana.)
In addition to being interactive, and loading the video behind the coding, Transana's maps and graphs can be configured, saved in JPG format, or printed.
Finally, creating clips and placing them in collections is rather one-dimensional, and limits the complexity of what can be inferred from a given clip. However, each clip can have multiple keywords associated with it, describing it along several dimensions. This allows for exploration of the coded data in much more sophisticated ways using Transana's search engine. Researchers can design sets of searches allowing them to ask very complex questions about their coded data and allowing them to explore subtle relationships between different aspects of the coding of their video data.