Computer-assisted qualitative data analysis: using NVivo

With your document open, you code your data by high-lighting a section of text and copying it to a node. There are many ways of highlighting text, like changing the font, or by using colour, which might be useful to you. NVivo simply accepts as nodes what you place there. Similarly, there are a variety of ways in which text selections can be inserted, including “dragging and “dropping” or by using the tool bar’s “copy and paste” tools. Again, it doesn’t matter to NVivo. It should follow from this that coding is absolutely your decision and any subsequent problems are more likely to stem from this fact, rather than from some defect in the programme. Since there are few practical restrictions on how much data can be coded, it is probably a good idea to code to a number of nodes at once. Later, redundant nodes can be deleted (or changed) and text can be “unselected”, if that seems more desirable.
Reference: Bryman: Social Research Methods: 5th Edition Page(s) 606-6111

CAQDAS is one of the most significant developments in qualitative research over the last twenty years. The label refers to a group of software packages such as NVivo and ATLAS/ti, which are used for qualitative data analysis. These computer programmes were developed for individual use on a PC or laptop computer and were originally not much more than sophisticated word-processors but have developed into advanced data-base programmes.
Reference: Bryman: Social Research Methods: 5th Edition Page(s) 602

Key concept 25.1 (p606) explains “nodes” as “collections of references” about aspects of your research data. In other words, nodes are like files in which you have inserted all references about a particular aspect of your research. NVivo has two types, called “hierarchical” nodes and “non-hierarchical” nodes. “Hierarchical” nodes are created by default, since they are for organizing your concepts in a hierarchy. By creating hierarchies of concepts you gradually start building towards a grounded theory. The “non-hierarchical” nodes are for what you regard as “free-standing” concepts – those that don’t seem to relate to other concepts directly.
Reference: Bryman: Social Research Methods: 5th Edition Page(s) 606,607

It is not only possible to create memos in NVivo, it is also desirable. “Memos” were discussed in chapter 24 as useful aids to the analysis of qualitative data. In NVivo, the sequence is as shown in answer (a), starting at the central, “navigation” screen. It is not necessary to write the memo first in another programme and then import it, since it can be done directly. This is one of NVivo’s strengths, in that it closely relates to the gradual, unfolding way in which qualitative analysis is typically done. Memos can be written during coding, or searching, for example, without disturbing the data array. Because memos are grouped separately they cannot be confused with other sources of data.
Reference: Bryman: Social Research Methods: 5th Edition Page(s) 615-616

Within the field of quantitative data analysis, SPSS is the most widely known and used statistical software package. It was developed many years ago, even before the advent of Windows and has been constantly improved and made more ‘user-friendly’ ever since. CAQDAS is a relatively recent arrival and depends on the ubiquity of personal computers. So far, no clear industry leader has emerged. However, most professional researchers pay close attention to developments in the NVivo software, particularly as it continues to allow manipulation of a variety of file types, including ‘media’ files. Broadly speaking, all CAQDAS programmes are to SPSS, as data-base software is to spreadsheets. NVivo was designed for Windows but can run on other operating systems with a bit of ingenuity.
Reference: Bryman: Social Research Methods: 5th Edition Page(s) 602

There are three main types of search that you can conduct in NVivo. If you want to see all data coded under a particular node (single node search), select the node in question at the “navigation” window. You can also search to find instances of text coded for two separate nodes, by using the “queries” button to open a “Coding Query” dialog box. Using the “Edit” button on the top tool bar, then “Find”, opens the “Find Content” dialog box (shown in Plate 25.11, p615), which helps you to search for all instances of a particular word or phrase. This type of “string” search may uncover “in vivo” codes (expressions used by research participants), which can then be created as nodes in their own right.
Reference: Bryman: Social Research Methods: 5th Edition Page(s) 611-615

The opening screen, shown in Plate 25.1 on page 604, is the “Welcome” screen. Selecting your project (or starting a new one) opens the central screen in NVivo10, referred to as the “navigation” screen. Having imported the project files that you want to analyze, you can open each one and edit it as if it were a Word document, using the Document Viewer (see Plate 25.3, p606). This is also where you can code your documents by applying nodes to sections of the data.
Reference: Bryman: Social Research Methods: 5th Edition Page(s) 604-606

There are numerous advantages to using CAQDAS, most of which centre on its speed and efficiency and the way in which mapping out “coding trees” of related ideas helps you to develop a grounded theory. While each CAQDAS program is unique and involves slightly different screens, functions and ways of representing the data, the basic techniques of importing, coding, retrieving and searching will be common to all of the programs and so provide you with a useful transferable skill. Obviously, new computer programmes have to be learnt and this takes time but the reward in this case is not just of skill acquisition but of helping to make the data processing and analysis transparent.
Reference: Bryman: Social Research Methods: 5th Edition Page(s) 603

The NVivo version currently available is NVivo10 and it is this version which is described in chapter 25. Earlier versions had a problem with importing documents other than text files but NVivo10 can import Microsoft Word documents (.doc and .docx files) directly, as well as .rtf and .pdf files. It is safer to convert other document types to “.doc” or “.docx” formats before importing them into NVivo10, because they are easier to work on later. In addition, many other non-text file formats can be imported and worked on, including a variety of audio and video files. The best advice is to check out which version is available to you and visit the website for further assistance (www.qsrinternational.com).
Reference: Bryman: Social Research Methods: 5th Edition Page(s) 604

Various criticisms have been levelled at CAQDAS, including the idea that it has created a new orthodoxy of “code-and-retrieve” qualitative analysis, that it fragments the data and that it is only useful for certain kinds of qualitative data. However, most researchers would agree that CAQDAS offers a faster and more efficient way of analyzing qualitative data than the alternative, which means doing the work manually. The risk of CAQDAS is that rich data can get lost in the computer and that we might come to see computer-generated output as being more “significant” somehow, than if we had performed all data processing ourselves.
Reference: Bryman: Social Research Methods: 5th Edition Page(s) 602,603