A Resource Guide for NVivo

IACC Project

April, 2005

 

This is a brief introductory overview and guide to NVivo for researchers and research assistants undertaking community vulnerability assessment. NVivo will be used for the analysis of the collected data and preparation of project reports.  The purpose of this guide is to introduce researchers to the NVivo software in a cursory manner.  The guide lays out some of the basic characteristics and functions of NVivo and it should be used with the guides developed by the makers of the software.

 

NVivo is a software program developed by QSR.   For an overview of the software’s capacities, see QCR’s NVivo website and the NVivo product brochure.  A very useful document is QCR’s Getting Started in NVivo, a guide that comes with the software but that can also be accessed on-line in pdf format at the above link. Similarly, the software comes with a more detailed reference guide.  The NVivo software comes with tutorials that can also be accessed on-line, which can help the researcher to further understand the capacities of the software.  Another, useful resource is a book entitled Qualitative Data Analysis: Explorations with NVivo written by Gramham Gibbs – this will be made available to researchers.  The NVivo software manual has been translated into Spanish for more information on Introducción a NVivo (2): Manual de Trabajo see the following link. 

 

Field methods for vulnerability assessment research undertaken in this project will take various forms including background histories, oral histories, structured and unstructured interviews and focus groups.  For more information on this approach see the Resource Guide for Community Vulnerability Assessment developed for the IACC project.  In out study of institutions and communities each form of observation will be recorded and transformed into computer files by the researcher.  The result will be a large body of texts – interview transcripts, field notes and other observational records – that comprise the ‘data’ to be organized and analyzed.  Qualitative data analysis consists of interpreting this body of data by organizing and drawing links between ideas and concepts emerging from the texts.  Researchers then may develop more general concepts and theories, while retaining its complexity, that help make sense of the data in light of the research questions driving the project.

 

NVivo facilitates several aspects of the qualitative research process by helping record information, organize and code texts, and develop systematic links between texts and emerging theoretical concepts.  The software allows researchers to record observations directly into Nivo or to import them into the program.  Once data have been imputed or imported, codes can be assigned to texts to begin to organize and manage the data.  Once we have created a database consisting of texts with their associated codes, researchers’ comments, and notes, the software allows for the easy organization and retrieval of portions of texts linked by common codes or concepts.  In turn, this allows the research to explore patterns in the data and begin to conceptualize the findings.  The data can be grouped and organized using different methods that assist the research in drawing out key findings.  The NVivo software includes various tools for displaying the systematic relationships between coded texts, for example, in ‘trees’ and ‘models’.

 

The first window to open when starting the NVivo software is the Launch Pad (Figure 1).  The Launch Pad provides four possibilities as well as a tool bar similar to windows.  From the Launch Pad you can:

 

  1. Create a Project
  2. Open an already existing Project
  3. Open a tutorial or
  4. Close the Program

 

FIGURE 1

 

Creating a Project

 

Even before researchers enter the field, background research has begun and so has the process of documentation.  Using NVivo, researchers begin by creating a Project in which different documents are created and organized into different categories.  For example, a researcher may have separate documents for background history, structured interviews, unstructured interviews, semi-structured interviews, and focus groups as part of a Project.  A Project is created by clicking on “Create a Project” from the Launch Pad screen. You do not need to have information in order to create a project. Documents could be added at any moment to a project that already exists. Projects could be “typical” - a researcher or a group of researchers could have access to the project without restrictions - or “customized’, where a member of the team acts as administrator and other researchers need a password to have access to the project. Once you choose to create a Project, the Project Pad screen is displayed, as shown below (Figure 2).  Existing projects are also open in this window by clicking in “Open a Project”

 

 

 

 

 

 

 

 

FIGURE 2

 

The Project Pad screen allows us to manage documents and nodes in NVivo.  Documents are different forms of recorded data created in the process of field research.  For example, you will have documents from interviews and focus groups (i.e. you could have as many documents as interviews and focus groups you had). Nodes are concepts, themes or categories that emerge from your documents.  The Project Pad contains two tabs, one for Documents and one for Nodes, both of which allow you to explore and manage your project.  The Project Pad also gives you access to the Attribute and Set tools.  Both Documents and Nodes can be given values of attributes and can be grouped into sets. 

 

Creating a Document

 

            Once you have started your project, information you collect while in the field will be saved and modified in different documents. The first tab on the project pad (Figure 2) is for documents.  From the Document Tab you can:

 

  1. Create a Project Document
  2. Explore Project Documents
  3. Browse a Document

 

To create a new document click on “Create Document”. From here a window with the New Document Wizard will open, using which allows you to create or to import documents as desired.  The Document Wizard will give you the following options:

 

  1. Locate and import readable external text file(s) that you want to use as document(s).
  2. Make Proxy Document(s) representing external file(s)
  3. Make a Proxy Document for non-file data
  4. Make a new blank document

 

If you choose “Explore Project Documents” from the Document Tab, the following Document Explorer window will open (Figure 3):

 

FIGURE 3

 

 

The tool bar along the top of the Document Explorer allows you to perform various functions including browse, properties, attributes, Doclinks and Nodlinks, which are be discussed in more detail below.  The window on the left of the Document Explorer allows you to view recently used nodes, open sets, drag and drop documents to make sets and to select documents and view size and coding status.  These functions will become clearer as each of these is described.  The window to the right provides details about each document including information such as the document title, size, nodes, creation date, etc.  For example, if you have one document for each community interview, you will see in this window as many documents as interviews you have. You can, of course, collapse all of part of the documents into a single one.

 

Finally the “Browse Documents” option allows the user to apply basic word processing functions such as changing the font size, font style, colour, etc. to a document.  This is done within the Document Explorer. 

 

It is important to note that in the case that data are not initially recorded in an NVivo document, they can be imported into NVivo from other formats such as Microsoft Word.  However, information to be imported must be saved as a Rich Text Format (.rtf) document.  You could also have complex documents, with datalinks to other types of files (audio, films, photographies, etc.).

 

Creating Nodes in the coding process

 

            Coding in NVivo is done through the use of nodes.  The QCR tutorial notes that, “nodes are containers for ideas and concepts and hold references to the passages of document text that you code at them”.  Coding is done by linking a passage in a document to a node. For example, you could create the node “drought” and code one or more interviews (one or more documents) using this node. You will end with a project node document that will contain all the sections of the interviews that are related to the event “drought”. You then could analyze this project node document looking for a more detailed account of the different understanding of “drought”.

 

            As we have seen, the first tab on the project pad is used for creating, exploring and browsing documents; the second tab is used for creating, exploring and browsing nodes.  In the initial stages of coding, your coding categories will be free nodes, meaning they will be unclassified and not part of a tree or case structure, which will evolve as your data analysis evolves.  You can view the coding you have completed in the Node Browser.

 

The Node Browser appears and functions much like the Document Browser.  It allows you to see document text that is coded at a specific node.  Like the Document Explorer, the Node Explorer (Figure 4) - has a tool bar with various functions including browse, properties, attributes, DocLinks, Nodlinks.  Unlike the Document Explorer, it also has the functions Edit Set, Assay and Search included in this tool bar.  The left window of the Node Explorer allows us to see recently used Nodes, the number of each type of Node, and sets.  The window on the right provides details such as document title, size, nodes, creation date, etc. 

 

FIGURE 4

 

 

The Node Tab on the Project Pad gives you the following options:

 

  1. Create a Node
  2. Explore Nodes
  3. Browse a Node

 

If you choose to create a node you will be given a choice between three types of nodes:

 

  1. Free Node
  2. Tree Node
  3. Case Node

 

Free Nodes are used when categories cannot be clearly separated into a system that the researcher wants to treat as independent categories.  These are often initial concepts or categories that emerge from the data, which need further development or classification. Free Nodes are nodes you have not arranged into relationships with other nodes.  If you choose to create a Free Node, a window opens in which you can give the Free Node a title and Description. 

 

            Unlike Free Nodes, Tree Nodes are nodes that have been organized into hierarchies of categories and subcategories, similar to the branches of a tree.  Tree Nodes are developed using the Node Explorer window. The QCR website defines a tree node as “a node which is organised in a tree structure. Node trees normally have fairly generic concepts at the root of the structure and progress upward to more specific concepts”.  For example, you could start using “Farmer” as a free node, but later you may decide to make distinctions among farmers, so that you develop “branches of farmer” as nodes: “grain farmer”, and “livestock farmer” or “mixed farmers”. Like Free Nodes, they are given a title and a description; however, they are also given an address so that they can be marked and ordered.  Finally, the third type of node is a case node.  A case node allows you to contain material about each case within this case node.  Like the previous example you may have a case type node called “Farmers” and then each individual case node would be created under it storing the material for each of the farms in the project. 

 

            While using nodes will probably be the most common way to code the data, visual coding or coding with attributes are also possible.  Further, coding through nodes, by visual codes or attributes can be done separately or can be used in combination.  Visual coding is done using the Document Explorer. 

 

            NVivo allows for more than one way to code.  You can code using the coding bar (Figure 5) found at the bottom of both the Document Browser and the Node Browser.  Here you select the text you want to code and the node you would like to code it to and click on “Code”.  Using the same procedure, you can uncode. 

 

FIGURE 5

 

 

Alternately, you can create a new node by selecting text and clicking on “In-Vivo”, which creates a new free node.  Also, from the Document Browser and the Node Browser you can click on Coder, which will open the Coder window.  From the Coder window you can select the text you want to code by either selecting the node you want to code it at and clicking “Code”, or by dragging and dropping the selected text onto the node.  Finally, you can code using the paragraph Coder, Section Coder or the Search Tool.  For more information on these please see page 32 in Getting Started in NVivo.    

 

Adding your ideas to the data

 

            As you read through your data, which will be organized into documents, you may wish to add some of your own ideas as the data is recorded.  This may help as data analysis is undertaken.  An annotation can be attached to text as a hyperlink, very much in the same way that you will write a note in the margin when reading a book.  This is done from the Documents Browser by selecting links and then attaching Databite.  To review an annotation after its creation, select the relevant links and click on Inspect Databite.

 

Exploring the Data

 

            The Explore Document function on the Project Pad (Figure 2) allows you to see and organize documents you have created.  Once a substantial amount of data has been created or imported into NVivo, you may wish to run searches of the data for common themes.  This is done by using the search button on the project pad.  Searches can be performed on texts, nodes, attribute values or a combination of these. For example, you could ask NVivo to search for all the interview documents that have the node “drought” and separate them by the gender of the respondent (an attribute).   After running a search, you can collect the documents found into a node and or attribute for further reference.

 

            While exploring the data you may wish to create connections between different elements within the data.  This can be done in three ways: using DataBites, DocLinks or NodeLinks.  DataBites connect a section of text with another section of text or an external data source.  DocLinks are used to connect a document, a node or a place in the text with any number of other documents.  Like DocLinks, Nodelinks connect a document, a node or place in the text with any number of other nodes.  These connections can be established, modified or removed as necessary.  For more information on linking data, see page 28-29 in Getting Started in NVivo. 

 

Creating Attributes

 

            An attribute is a property (for example the age and /or gender of the respondent the day and/or time of the interview, or any other feature that you think is important) assigned to a document or node that allows for further classification and organization of the data.  From the Project Pad, select Document then Create/Edit Document attribute.  Next, type the name of the attribute you want to use and click Apply.  Attributes can be viewed and modified in the Attribute Explorer.  An attribute is most commonly a demographic variable when the document represents the ideas of an individual.  For example, if you wanted to look at how different generations answered one of your interview questions, you might include age as an attribute.  You may also employ a null value such as Not Assigned (-), Unknown (?), or Not Applicable (NA).

 

The following figure (Figure 6) shows us how to develop and define attributes.  The window of the left lists existing attributes and the right screen allow you to create a new attribute and/or to edite a selected attribute.  Each new attribute which is created is given a name, description, value type and, if desired, a default value.  The selection of a value type is facilitated through a dropdown window which supplies attribute values such as strings, Boolean (true/false) numbers, dates, times etc.  The default value is unassigned, that is represented as (-). 

 

 

FIGURE 6

 

 

Creating a Set

 

Documents can be grouped into sets for further administrative organization.  You may want to group all of the documents from a certain data collection period, or from one respondent or type of respondent, for example.  For instance, we may want to establish a set of documents of interview with just “business people” – we can define this as a specific set allowing us to focus analysis only on this type of community member.  A set is a grouping of documents or nodes you can use in order to access and work with these data as a group.  Sets are used to assist in coding, to help develop models, to perform searchers, to group results from analysis, to group memos by themes, or to group nodes belonging to a specific area of analysis or to further organize the data. 

 

Sets can be created using the Document Browser or the Node Browser.  From the Project Pad select either or these and from there you can create or modify sets.  To further edit the sets that you have created, you can use the Set Editor. The Set Editor will allow you to use the properties of a document to select them for use in sets.

 

For information on filtering a set or clarification on creating sets see page 34 in Getting Started in NVivo. 

 

Creating Case Nodes and Node Attributes

 

            To further organize and understand the data collected, you may want to create case nodes.  For example, each participant interviewed in a focus group could become a case.  A case node is created in the Node Explorer.  We have seen how to create document attributes to organize documents; similarly, node attributes can be created to sort demographic information.

 

Conceptualization of findings

 

            As previously discussed, some of the ways the researcher can conceptualize project findings is by using the Document Explorer to show relationships between documents and by further organizing them into sets.  The Node Explorer also allows us to examine different nodes and relationships between them.  Furthermore, nodes can be organized into sets using a Tree Node or grouped into Sets. 

 

            Besides sets, another way to show relations in your data is to use the Show Tool.  From you Project Pad, select “Analysis” and then “Show Tool”.  Once you select the items you want to look at, an explorer window will open from which you can use the item menu to carry out functions such as browse, edit, and explore the nodes, documents, attributes and sets as well as removing them if desired.  Like the Show Tool, the Assay Tool is used to show relationships in your data; however, the Assay Tool is more detailed than the show tool.  For instance, it can be used to show the proportion of documents in a set coded at a particular node.  From the Document or Node Explorer, select Document and then Assay Document.   The Assay Tool also allows you to develop a table to show your results.  For more information on the Show Tool and the Assay Tool see page 36 and 37 in Getting Started in NVivo. 

 

Using a Modeller is another way NVivo allows us to conceptualize the data.  A Modeller uses icons to represent a range of features of the data and allows us to establish relationships between icons.  To begin a model, open the Model Explorer by clicking on Explore Models from the Project Pad.  Subsequent items nodes, links, properties, labels, and subtrees can be added to the model.   Models can be layered as the process of data analysis develops.

Searching the Data

 

            Another helpful tool is the Search Tool, which will allow you to search the data and if necessary group items found in the search.  From the Project pad, click on Search Project Database and the Search Tool window will open.  The Search Tool window allows you to:

 

  1. Run the search operation, allowing you to define what you want to search
  2. Select the scope or your search
  3. Decide what you want done with the search results

 

In defining what you want to search you can choose to search a node, an attribute or text or more than one of these if desired.  Simple searches are done when you search a single dimension such as a node, attribute or text.  To search a node you can also use the Node Explorer.  Similarly, to search a value in an attribute you can use the Attribute Explorer. 

 

An example of a simple search:

 

To search a specific node, click on the Node Lookup icon.  From here you will need to indicate which node you want to search.

 

You can also run a simple search of unique attributes or of text – a word or phrase. 

 

            Besides simple searches, NVivo can run relational searches such as Boolean and Proximity searches.  Unlike a simple search, relational searches deal with more than one node, attribute or text at a time.  Boolean searches use commands such as “and”, “or” “not” and “less” to generate more complex searches, whereas proximity searches seek items that are near, precede or surround other items.  For more information on the six different types of Boolean searches (intersection, union, negation, difference, matrix intersection and matrix difference) see page 45 in the Getting Started in NVivo.  Similarly, for more information on Proximity searches (co-occurrence, sequence, inclusion, matrix co-occurrence, matrix sequence and matrix including) see page 46 in the guide.

 

Conclusion

 

            The researchers and research assistants involved in the vulnerability assessment of communities and research on institutional capacities to deal with such vulnerabilities will be provided with a copy of NVivo Version 2, as well as the accompanying manuals, which can also be accessed on the internet (the links are provided in the introduction).  Researchers will also be provided with the QSR Merge for NVivo program which is allows teams of researchers, such as those working on the IACC project to merge data together for further analysis.  For an over view please see QSR Merge for NVivo. 

            Finally, QCR’s website provides access to different support mediums should the researcher encounter difficulties while using NVivo.  QCR’s website offers a Help Desk in which help via e-mail correspondence can be obtained.  Similarly, the website offers Forums in which product users can post inquires, search past inquires and communicate with each other.  

The Information Technologies Services at the London School of Economics offers the following brief guide:  Getting Started with NVivo 2.  As well, Patsy Clarke of the University of KwaZulu-Natal has developed the following Click-State Guide: NVivo version 2, which is another helpful resource guide.  Sandrine Zerbib has developed a Learner’s Guide to NVivo 1.2, which provides step by steps links to the various functions on NVivo.  While this guide was developed as a tutorial to NVivo version 1.2, many of the functions remain the same in the updated 2.0 version and therefore much of the guide remains helpful and relevant.  For specific information on preparing your field notes for Nivo, see Preparing text documents for NVivo. 

 

Hamilton Jemmott has written an article entitled Using NVivo in Qualitative Data Analysis, which describes her experience in using NVivo in her research which is example of how the software is used. Similarly, Elaine Welsh’s article Dealing with Data: Using NVivo in the Qualitative Data Analysis Process accesses the way in which NVivo is used in the data analysis process. Anette M. DeNardo and Lisa Lopez Levers have written a paper entitled Using NVivo to Analyze Qualitative Data, which details their research experience using NVivo. 

For assistance regarding the use of NVivo for the IACC project please contact the IACC Project Director:

Polo Diaz

Department of Sociology and Social Studies

University of Regina

Regina, Saskatchewan

S4S 0A2

Phone: (306) 585 - 4151

Fax:   (306) 585 - 4815

E-mail harry.diaz@uregina.ca