In the introductory chapter, we defined a number of key terms that we will refer to throughout the guide. However, since the labels and terms in the coding software often differ, and unfortunately sometimes overlap, there is some risk of confusion when it comes to the core terminology of STCA.
In this note, we will map the core STCA terms (developed in the context of STCA or adopted from related network methodologies), their description in the context of the method, and the corresponding terms, functions or other related information in the software used (currently NVivo and Visone).
Mapped- and associating variables
In the introduction, we refer to mapped- and associating variables and use the actor and concept codes as general examples of such variables throughout the guide. We also refer to the process of coding at different codes as a core step in STCA. In other words, both associating- (for example actors) and mapped (for example concepts) variables are represented by codes.
In NVivo, the possibility to code text is a core feature of the software used in STCA. When coding in NVivo, however, you can choose between coding text snippets as codes or as cases. This discrepancy is a potential source of terminological confusion.
In Chapter 3, we present compelling reasons for why cases are to be preferred over codes. This means that when we are talking about coding and codes in the guide, this refers to using the feature coding as cases in NVivo. A code in the general STCA-terminology thus corresponds to a case in NVivo, if following the step-by-step instructions outline in the chapters of this guide.
Codes can in other words represent both mapped and associating variables, depending on their position on rows versus columns in the data export from NVivo.
In Visone, individual codes are represented by nodes in a network visualization, which are linked together by edges derived from the analysis of the data.
Partitioning and complementary variables
The introduction refers to partitioning- and complementary variables and the remainder of the guide provides step-by-step instructions of how to practically go about creating and using such variables.
In NVivo, we use the function of classification to assign meta-level information about the characteristics of a particular code (see Chapter 3). For example, an actor-code can be assigned information about the type, size and location of the actor. We also refer to these interchangeably as attributes. Attributes are stored in classification sheets.
The difference between partitioning- and complementary variables thus lies in how they are used in the context of STCA, not in how they are created or structured during the coding procedure in NVivo.
Typically, a partitioning variable is used during the data-export step in NVivo, to create sub-sets of the full database corresponding only to a particular time period, location, or similar. A complementary variable is stored in classification sheets and exported separately, to be used later in Visone.
We allude to the point that there are other ways of storing information that can be used as partitioning- and complementary variables, for example as codes or by manually creating attribute lists. This could be explored further by anyone interested.
In Visone, complementary variables are used to improve the visualization of a network by, for example, colouring actor-nodes according to type or size. This is done in Visone using the function of attributes and thus importing the attribute lists that you can export from NVivo (corresponding to the classification sheets in NVivo).