Think of a node as a container for qualitative data (themes, concepts). It is a way to tag your data and sources; it may appear that you are moving your data from its original source, but really you are simply adding tags to the source.
Before you even bring sources into NVivo it is worthwhile to consider creating separate sub-folders within the "Nodes" folder, for instance having folders such as "Auto Coded", "Query Output", "Themes", and "Cases".
- An "Auto Coded" sub-folder can be useful for storing any nodes that you allowed NVivo to auto-code, such as the different headings in a Word Document survey or questionnaire.
- A "Query Output" sub-folder can help to separate what you have mined from the data via queries.
- A "Themes" sub-folder can store the nodes that you have carefully extracted yourself from the data or sources.
- A "Cases" sub-folder can be useful for storing case nodes, or nodes that contain demographic data, for instance individual people can be considered as case nodes. You can find more information on creating Case Nodes in the Node Classification section of the "Classifications" webpage.
Creating nodes is called coding. Through coding you can look at the positioning of nodes (such as conflict and cooperation) and see if they are ever mentioned together in the same source and if so, which is mentioned first, and then explore what this could possibly mean. This sort of evaluation will allow you to see trends in your data.
There are three main ways to code your data:
2. Auto Coding
3. Query Coding
In this section you will also find information on:
When you are coding themes by carefully reading the contents of your sources, this is considered manual coding. The nodes created via manual coding should go in your “Themes” sub-folder.
Manual coding will allow you to see how two different themes relate to one another. For instance, if you want to know how many people feel positive about a topic such as coming home from college, versus how many people feel negatively about coming home from college, you can create three separate nodes, (positive, negative, and coming home from college). Eventually this will allow you to search the intersection of these nodes, and investigate how many people are speaking in a positive light and coming home from college, and what they are saying about the two themes.
You might be tempted to only create two nodes, one for positive statements about coming home from college and one for negative statements about coming home from college, but this will limit your coding potential. For instance, one of your informants may have mentioned positive feelings about coming home to see family in the same interview answer. If you had created nodes for “positive”, “coming home from college”, and “seeing family” you can search the intersection of all three of these nodes and discover whether or not “seeing family” influenced the positive feelings about “coming home from college”.
To Manually Create a Node:
Right-click on the appropriate parent folder in the upper portion of the Navigation View -> Select create new node -> A window will pop up with two tabs -> In the first tab, name the node, and type in a description and a nickname if you so desire -> Here you also have the opportunity to give the node a coding color and to aggregate the node -> On the second tab you can add a classification if necessary (for a person or institution).
Sometimes when reading a source you may realize that you wish to tag a section of it as a new node. In this case:
Highlight the text -> Right-click -> Select the “code selection” option -> Select “code selection at new node” -> A window will pop up which will allow you to select the folder you wish to place the node in -> Select the appropriate folder -> Click “Ok” -> Name the node and give it a description if you desire.
In order to give this new node a color, classification, or to aggregate it you will need to open the “Nodes” folder from the Navigation View -> Locate the node -> Right-click -> Select “Node Properties” from the drop-down menu and alter the necessary information -> Click “Apply” -> Click “Ok”.
Adding Data to an Existing Node:
Highlight the portion of text -> Right-click -> Select the “code selection” option -> Select “code selection at existing node” -> Select the appropriate parent folder to find the node you want to code the selection at -> Check the box for that node -> Click “Ok”.
If you have already tagged other data at that node during your session, you may see it below the “code selection at existing node" on the drop down menu. This is a shortcut for tagging at an existing node. Simply choose that node from the drop-down menu.
You can auto code your data if you have imported documents with Heading Levels.
In order to auto-code:
Select the files you want -> Right-click -> Select auto code -> Select which heading level you want to separate out, such as interview questions -> Move to the next section via the arrows -> Select the folder you want that node folder to come up in -> Name the node -> Click “Ok”
*This is useful if you want to look at all the responses to each interview question and see the responses all together.
Adding Data to an Existing Auto-Coded Node:
Highlight the desired portion of text -> Right-click -> Select the “code selection” option -> Select “code selection at existing node” -> Click on the appropriate node folder to find the node you want to code the selection at -> Check the box for that node -> Click “Ok”
If you have already tagged other data at that node during your session, you may see it below the “code selection at existing node” and “code selection at new node” on the drop down menu. This is a shortcut for tagging at an existing node. Simply choose that node from the drop-down menu.
Whenever you run a query, you have the option of turning that query into a node. Placing this node into the “Query Output” folder will help you keep the nodes created by mining the data via queries from the data you carefully extracted via manual coding.
In order to see the query results double-click the appropriate node, double-clicking on the query will cause the query to run again.
To add another piece of data to a “Query Output” node:
Highlight the desired portion of text -> Right-click -> Select “code selection” -> Select “code selection at existing node” -> Click on the appropriate node folder to find the node you want to code the selection at -> Check the box for that node -> Click “Ok”If you have already tagged other data at that node during your session, you may see it below the “code selection at existing node” and “code selection at new node” on the drop down menu. This is a shortcut for tagging at an existing node. Simply choose that node from the drop-down menu.
If you make a mistake while coding you can uncode at existing node or existing nodes depending on how many nodes you wish to erase the information from.
In order to uncode at existing nodes:
Simply highlight that portion of the source -> Right-click -> Select "Uncode Selection" -> Select "Uncode Selection at Existing Nodes" -> Find the node you wish to uncode at -> Uncheck the box to uncode -> Select "Okay".
Or uncode by:
Select the "Analyze" tab from the top menu ribbon -> Select "Uncode Sources At Exisiting Nodes" -> Find the node you wish to uncode the source at -> Uncheck the box to uncode -> Select "Okay".
Uncoding By Coding Stripe:
If you are using coding stripes and wish to uncode: Right-click on the coding stripe you wish to uncode at -> Select "Uncode".
- Be wary of the "Uncode selection at current nodes" option. This is the most difficult to use, as it will uncode the data from which ever was the most recently used node aka the node you can see in the lower ribbon on the main page, but if you happened to click on a new node by accident or before you uncode the selection, you run the risk of uncoding at the wrong node.
It is important to carefully consider the label of each node. You want the name to be something recognizable to you and any partners you are working with, as well as a proper reflection of the sort of data you are tagging with it.
When you make a new node that is where you can make a nickname, but there is not much benefit to the nicknames unless you are having difficulty remembering the name of your node.
You can also color code if you wish, which may be more useful if you are working with another partner, or if you are the type of person who color codes their data. If you do not code a color your coding strips will fall back on an automatic color scheme.
You can also add a description to certain nodes if you wish.
You have the option of organizing your nodes into hierarchal relationships. For instance if you have a lot of nodes that are unique dates, you may want to create a parent node to organize the dates by year and by month.
It is important to consider what you choose to label the nodes, for instance labeling numerically (i.e. yyyy-mm-dd) will ensure that the dates are in proper order.
Create a Node Hierarchy:
- In order to move individual nodes into parent nodes -> Highlight the node(s) you wish to move -> Right-click -> Select “cut” from the drop-down menu -> Right-click over the parent node -> Select “paste” from the drop-down menu.
Example of a Node Hierarchy:
Parent Node -> Attitude, Child Node-> Positive, Child Node-> Negative
Aggregating allows you to suck all the child nodes into the parent node. Once you aggregate a parent node, it will pull up any new data you add to any of its existing child nodes or any new child nodes that you add.
- To aggregate a parent node: Right-Click on the parent node -> Check the aggregate option.
- You can change this option at any time. To unaggregate a parent node: Right click on the parent node -> Uncheck the aggregate option.
Coding Stripes can be extremely useful to you NVivo project. They can allow you to easily view what nodes have already been tagged on a source.
- Click on the “View” tab on the top bar -> Click on “Coding Strips” -> Select the appropriate option from the drop-down menu. Bear in mind, the “Nodes Recently Coded” or “Nodes Most Coded” will show the most coding stripes.
- You can also set the coding stripes to appear automatically for every source. In order to do this: Go to “File” -> Select “Options” -> A new window will appear -> Click on the “Display” tab -> Under “Coding Stripes” select the appropriate option -> Directly below this option you can set the maximum number of coding stripes you want to see at a time -> Choose your settings -> Select “apply” -> Select “okay”.
In order for this option to take effect, it is necessary to close your project and re-open it.
Uses of Coding Stripes:
- You can click on the coding stripes and it will highlight what was coded.
- You can right-click on a coding stripe and uncode the highlighted information from the node.
- You can also open a node by right-clicking on the coding stripe and selecting "open node".
- Coding density is another stripe you can use and it will show you the part of the text/data that you have coded the most in a certain source. You can hover over the dark patches and a bubble will appear that will allow you to see what has been coded at that area.
When you open a node (by clicking on the desired node in the upper portion of the Navigation View) you will be able to see all of the information coded under that node. Sometimes you may find it useful to see more of the context surrounding the coded selection.
In order to change the coding context:
Highlight the node -> Right-click -> Select “Coding Context” -> Select:
- “None” (for no context)
- “Narrow” (which is five words)
- “Broad” (which will be the full paragraph)
- “Custom” (which you can set)
- “Entire Source” (which will pull up the entire source quickly)
The context that appears will be the color grey to indicate that it is temporary. In order to get rid of any coding context, right-click and select “none”.
In this folder you will be able to find, open, and edit any relationships that you have created. A Relationship is a certain type of node used to define the relationship between two different items of your project, such as how one node affects another node or how a certain case affects a certain theme, etc.
Relationships allow you to investigate theories you may have about these different items. For example, if you think there is a relationship between sunlight and flower growth, you can examine that theory by creating a relationship such as sunlight 'impacts' flower growth.
A Relationship acts like a node in that you can code selections of text into a relationship. Creating a new relationship is very similar to creating a new node.
Create a New Relationship:
Right-click on the Relationship folder from the upper portion of the Navigation View -> Select "create new relationship" -> Connect the two items you want to put in a relationship by selecting an item for each the "From" and "To" section -> Select the relationship type-> Select "Ok".
For instance the relation type for "sunlight 'impacts' flower growth" would be a one-way relationship. For more on types of relationships, refer to the "Relationship Types" section of the "Classifications" webpage.
To add a selection to an existing relationship:
Highlight the desired portion of text -> Right-click -> Select the relationship you wish to code the selection at.
In this folder you will find any of the Matrices that you created. You will also be able to view their properties, edit them, create a set from them, and export them by simply right-clicking on the Matrix and choosing the appropriate action from the drop-down menu.
A matrix is a grid that provides a way to summarize source materials, nodes, sets, or other data by case and theme. Matrices can also help you to get familiar with your data, see patterns and overlaps, such as how different themes relate to one another or to a particular case, person, or institution. It is also useful for sifting through a vast amount of information and making easy comparisons.
- You can compare the nodes such as "small businesses" vs "large businesses".
- You can compare survey answers. In "nodes coded" choose a demographic node such as “person” or “institution” in order to see how many people used a certain word or are connected to a certain node in their data.
Creating a Matrix:
Node matrices are created by running a Matrix Coding Query. Directions on how to run a Matrix Coding Query can be found on the Query web-page.
- After you create a matrix, you can click on the data and change the context.
- You can also right-click on the cell or below the cell, choose “cell content” from the drop-down menu, and then change what the table is showing you. For instance you can change from “coding references” to something like “nodes coded”.
Sharing a Matrix:
You can easily share this summary of data with others by printing or exporting the matrix:
Right-click the matrix -> Select the export or print option.