The index system (‘Node Explorer’)
This system provides a tool thinking about your data, and for developing and exploring ideas about the data. Gahan and Hannibal (1998, p. 9) view the document and index systems as two separate metaphorical bags sitting alongside each other. The document bag contains your data in a variety of documents which are mostly complex, messy and pretty chaotic. The index bag has nothing in it to start with. You put into it things which appear to you to be likely to help in analyzing and understanding your project. These include categories, themes and labels which express characteristics of things in the data; your first attempts at developing theories about what is going on. In other terms, your coding of the data goes into the index bag.
Coding in NUD*IST is the process of indexing your text so that the coding scheme can be used as an index for retrieving text. The codes are (somewhat confusingly-though you get used to it) referred to as nodes. This is because NUD*IST encourages you to develop a hierarchical ‘tree’ structure, where the nodes are the points in the tree where branching takes place, as illustrated in figure 14.1. It is, however, up to you how the nodes are organized. They can be entirely unstructured, or be in a flat non-hierarchical structure; or some may be left free-floating while others are linked in a tree. As you proceed with the analysis and exploration of the data, it is likely that at least some degree of structure will develop. You can construct a node tree without indexing or coding any text at the nodes, for example if you have a pre-existing template to use.
It is worth stressing that nodes can represent anything you want. For example, as well as categories or concepts they can refer to individuals, or specific contexts, or to a case. As discussed below, one node in a grounded theory study might concern all the demographic information about those interviewed; or, in an ethnographic study, a node might relate to descriptive information about the culture being studied. A particular text unit can be coded in more than one way.
Whenever you create a new node, NUD*IST will ask you to specify its position in the index tree hierarchy, or to create it as a free node. If, as is likely at the initial stages, you are not sure about this or haven’t thought about it, don’t let the program bully you. You can leave a node, or all nodes for that matter, as free nodes for the time being, and assign them a place in a developing hierarchy at a later stage