Understanding Auto-Coding in Nvivo

Earlier we discussed creating nodes that act as ‘sticky notes’ denoting particular themes or topics common across the qualitative data. However, the earlier article was partial in information since it dealt only with manual coding of themes and elements or topics into nodes. Nvivo, both pro and plus provide an additional feature of auto-coding documents. While working with structured documents like interviews where the same set of questions are asked to different participants, Nvivo allows the documents to get automatically coded in respective nodes. This saves time from going through the whole data (especially transcripts) in case the sampled population is huge. However, to initiate auto-coding in Nvivo, it is necessary to confirm the document is well structured and designed in a prescribed format. Therefore, a step by step guide has been provided to help utilize the auto-coding option.

Understanding auto-coding in document sources

Nvivo provides ways to auto-code documents. For instance, in the case research, we have imported eight-word files as interview transcripts in Nvivo. Now, we can auto-code the word documents based on paragraph style. Paragraph Style allows organizing the document in heading styles. Organizing transcripts as per different heading styles assists Nvivo in identifying and auto-coding the content. For example, an interview transcript of Preeti is shown in figure 1.

Here, the transcript contains the responses provided by Preeti during the interview. However, the document has not been organized as per any heading style. For auto-coding, we can format the interview questions into a specific heading style (e.g. Heading 1) and answer in ‘Normal’ font before importing the documents.

Since it may happen that the transcripts have already been imported without formatting but they still can be formatted in Nvivo. To edit the document in Nvivo, follow these steps:

1. Click on Sources
2. interview (in case of transcripts)
3. double click (Interview Preeti in case research)
4. the transcript will open in detail view as shown in the above figure

To edit the transcript, click on the click to edit icon on the top of the document (as shown in the above figure). Now to format the document within Nvivo, choose the content and with the help of editing tools in “Home” (in the ribbon), apply the style. For instance, select the heading of the transcript, as in figure 3, and select “Title” form the style option in the ribbon.

Similarly, we can edit all the questions one by one in the same way. For example,

1. Select question 1
2. Click on the Home icon (Ribbon)
3. go to style options
4. select heading 1 

Similarly, apply heading 1 in all the questions and making similar changes in all the other transcripts. In case, if the transcripts have sub-questions as well, then we can assign heading 2 to the sub-questions for Nvivo to identify and differentiate them from the main question. For example, the transcript of Preeti also has a sub-question as shown in figure 5.

1. Select question *2a
2. Click on the Home icon (Ribbon)
3. Go to style options
4. Select heading 2

Similarly, all the questions in the transcript can be organized through changing style.

To start with auto-coding

After all the transcripts are formatted in Nvivo, we can auto-code the transcripts:

1. Click on interviews (Sources)
2. Click on Interviews
3. Select all the transcripts at once (press CTRL)
4. Click on ANALYZE (in the ribbon)
5. Click on Auto Coding

A dialogue box named “Welcome to Auto-Coding Wizard” will open in Nvivo Window as shown in figure 6.

First of all, specify how to code the content of Transcripts. For that select “Paragraph Style” in the Dialogue box as shown in the above figure. After selecting “Paragraph style”, select the type of paragraph style as per the organized transcript. For instance, in the case research, the main questions in transcripts have been organized as per heading 1, click on heading 1 and through the rightward arrow, put heading 1 in the selected paragraph style box. Further, press next as shown in the below figure.

After clicking on Next, step 2 of the dialogue box will appear as shown in the below figure. Step 2 contains three options, “Under”, “Location” and “Name”.

In the drop-down option “Under”, select “Existing” or “New Node”. If Nodes have already been prepared then manually select “Existing Node”. If not created then select “New Node”.

While selecting the “Existing Node” option, select the location of an existing node in “Location”. While choosing “New Node”, the location option comes up with “Nodes Folder”. Lastly, give an appropriate name to auto-coding. For instance, in the case research, we have named the node “Sorted by Questions”. Press finish as shown in the below figure.

Viewing Auto-Coded Nodes

After finishing the procedure, transcripts in the Interview folder have been filled with a number of Nodes and References as shown in the below figure. The numbers in the “Node” column indicates the number of nodes coded with a particular transcript. The number of the reference column suggests the number of sentence codes in nodes from a particular transcript.

For more clarity, view the list of Nodes that have been made through the Auto Coding option.

1. Click on “Nodes” in “Main Menu”.
2. Click on “Nodes”.

You can see the “Nodes” folder “Sorted by Question” that we have made through auto coding. Furthermore, Question wise Nodes can also be viewed by clicking on the “+” sign, as shown in the below figure.

On clicking the “+” sign, the Nvivo window will come up with the list of nodes that have been created and coded as per the questions.

Therefore, just like manual coding different nodes are created based on the questions manually. Auto-coding assists in making nodes automatically through formatted transcripts. However, despite its ease of using and generating automated nodes, the option is not without its issues and limitations, which has been reflected in details, separately. Having discussed in detail two approaches of data processing the next article moves to understand memos, which enables recording ideas, insights and inferences related to the data and issue of the project.