Categorization and Coding

by | Analysis, Coding, Collections, Keywords

Transana offers two overlapping systems for indicating analytic meaning in text and media data, Categorization and  Coding.  In Transana, you code by creating a system of Keywords and applying these keywords to selections of text called Quotes or segments of media data called Clips.  You categorize data selections by creating Quotes and Clips in containers called Collections.

Both categorization and coding serve the same purpose, to link analytic meaning to segments of data, but they involve somewhat different metaphors and procedures within the software.  There are several reasons behind the evolution of the two systems for attaching analytic meaning to segments of data in Transana.

First, different qualitative methodologies have different requirements.  In Narrative analysis, for example, the order of items within a category is crucial, as the researcher orders the segments so they tell the story that emerges from the data in a coherent, organized way.  Grounded theory has little use for the concept of sequence within a category.

Second, different researchers prefer to work with data in different ways.  Those who learned “old-school” manual qualitative analysis involving making piles of slips of paper to represent different analytic categories may find that putting Quotes and Clips in categorical Collections feels more comfortable, while researchers used to querying databases may find coding to be more familiar in the way they are used to working with data.  These different tools match different analytic styles.

Finally, different systems may work better at different stages of the analytic process.  When I do an analysis myself, I often start by doing a lot of pure coding (making Quick Quotes and Quick Clips  Transana based on Keywords), which I eventually move into Collections as I formulate an understanding of what my data has to tell me.  Later in analysis, as my theory that helps me understand my data evolves, I do much more categorization (creating Standard Quotes and Standard Clips in Collections) to fit new segments of data into the existing or expanding theoretical structures that Collections represent.

Let’s explore the relative advantages of each system and how to move between them.

Among the advantages of categorization are that the data segments within categories are visible, concrete, and immediate. You can unfold the nested Collections and see how much evidence (in the form of Quotes, Snapshots, and Clips) you have for each of the theoretical constructs represented by your Collections. You can easily watch all the video clips, see all the coded images, and read all the relevant text gathered into each Collection. In addition, items within a Collection are orderable so you can use them to tell a coherent story with beginning, middle, and end. A disadvantage is that you can really only say one thing about a data segment by placing it in a Collection, the same as putting one slip of paper in one pile.

The advantages of Coding mostly center on flexibility. You can apply multiple codes to the same data segment to describe that segment along multiple dimensions. This opens up a variety of interpretive possibilities, including being able to create coding maps which show the relationships between different codes, and being able to craft searches which return subsets of your data to allow finer-grained exploration of what you have coded. A disadvantage of coding is that you really need to use Transana’s reports, maps, graphs, and searches to find the larger meaning that coding represents in your data because you can’t “see” coding across items as easily as you can see categorization. Fortunately, Transana’s reports make this reasonably easy.

It is important to emphasize that researchers can do complete analyses using either of these systems exclusively to analyze their data, and will likely reach the same conclusions either way. It’s also quite easy, once you understand the strengths of both systems and the tools they use, to combine the two approaches for the best of both worlds. Some methodologies (such as Narrative analysis) favor one approach, while other methodologies (such as Grounded Theory) can be accomplished equally well with either approach.

It’s also important to emphasize that there is less difference between these approaches than it might initially appear. You can easily add codes to a data segment in a Collection. You can easily create Collections and nested Collections to represent the most important ideas that emerge from coding. Since all data items can be both coded and categorized, it’s easy to combine both systems and move back and forth between them as needed.

While the dual systems for assigning meaning to data are sometimes a source of confusion for researchers new to Transana, it doesn’t take long to experiment with both approaches and settle in on the analytic process that works best for you, given your chosen qualitative methodology, your personal preferences, and your current analytic needs.