Coding Qualitative Data and Introduction to NVivo

What is Coding in Qualitative Research?

Coding is the process of assigning short labels (codes) to fragments of qualitative data (interview texts, field notes, documents) with the purpose of systematizing, organizing, and subsequently interpreting them. Coding is a fundamental stage of qualitative analysis and serves as a “bridge” between data collection and drawing conclusions.

According to Saldana (Saldana, 2021), a code is “a word or short phrase that symbolically assigns a summative, salient, essence-capturing, and/or evocative attribute to a portion of linguistic or visual data.”

Coding performs several functions:

• Data reduction — condensing large volumes of text into manageable units
• Organization — grouping related fragments for comparative analysis
• Conceptualization — shifting from description to analytical categories
• Ensuring transparency — documenting the analytical process

Types of Codes

Descriptive Codes

Descriptive codes summarize the topic of a data fragment in one word or phrase. They answer the question: “What is this fragment about?”

• Fragment: “Every morning I spend about an hour responding to emails before I can get to my main work.”
• Code: EMAIL MANAGEMENT

In Vivo Codes

In vivo codes use the exact words of the participants themselves. They preserve the respondent’s “voice” and are especially useful for capturing unique expressions.

• Fragment: “Here we’re all just ‘putting out fires’ every day, there’s no strategic thinking.”
• Code: “PUTTING OUT FIRES”

Process Codes

Process codes use gerunds (-ing, -tion, -ment) to denote actions and processes. They capture the dynamics of phenomena.

• Fragment: “At first, I tried to control everything myself, then I gradually started trusting the team...”
• Code: DELEGATING RESPONSIBILITY

Emotion Codes

Emotion codes capture feelings and emotions, either expressed or implied by participants.

• Fragment: “When they told me about the reorganization, I felt like the ground was slipping away under my feet.”
• Code: FEAR OF UNCERTAINTY

Values Codes

Values codes reflect participants’ values, beliefs, and orientations—what they consider important.

• Fragment: “What matters most to me is fairness on the team, so everyone gets equal recognition.”
• Code: VALUE OF FAIRNESS

The Initial (Open) Coding Process: Step-by-Step Guide

Step 1. Preparing the Data Ensure that all interviews are fully transcribed, numbered, and ready for analysis. Each document should have a unique identifier (e.g., INT-01, INT-02).

Step 2. First Reading — Immersion in the Data Read the entire dataset in full, without coding. The goal is to get an overall sense of the content. Make margin notes about first impressions and emerging ideas.

Step 3. Line-by-Line Coding Return to the start and begin assigning codes to every significant fragment. At this stage, code “generously”—it is better to create more codes than to miss something important.

Step 4. Comparing and Refining After coding 3–4 interviews, review the codes you have created. Merge duplicate codes, refine wording. Create a codebook—a table with the definition for each code.

Step 5. Continuing Coding Continue coding the remaining data, applying existing codes and creating new ones as needed. Periodically update the codebook.

Rules for Effective Code Naming

From Codes to Categories

After completing initial coding, the researcher moves on to focused (axial) coding—the process of grouping codes into more abstract categories and identifying connections between them.

Example of moving from codes to categories:

First-level codes:

• Lack of time
• Task overload
• Working on weekends
• Inability to disconnect
• Physical exhaustion

Category: PROFESSIONAL BURNOUT

The process of forming categories includes:

1. Sorting all codes by thematic similarity
2. Identifying key themes and patterns
3. Formulating category names at a higher level of abstraction
4. Checking categories for completeness—whether all codes have found their place
5. Identifying relationships and hierarchies among categories

Introduction to NVivo

NVivo is specialized software for qualitative data analysis (CAQDAS — Computer-Assisted Qualitative Data Analysis Software), developed by Lumivero (formerly QSR International). NVivo helps organize, code, and analyze unstructured data: text documents, audio, video, images.

It is important to understand that NVivo does not analyze data automatically–the software is a tool that helps the researcher systematize and manage the analysis process. All analytical decisions are made by the researcher.

Main Components of the NVivo Interface

Sources

This is the section where all your data is stored: interview transcripts, field notes, documents, audio and video files. Sources can be organized into folders by data type or collection stage.

• Documents — text documents (Word, PDF, text files)
• Audio/Video — multimedia files
• Memos — analytical and reflexive notes by the researcher

Nodes (Codes)

Nodes are the equivalent of codes in NVivo. Each node represents a theme, concept, or category. Nodes can be organized hierarchically—parent nodes (categories) contain child nodes (subcodes).

• For example: the parent node “Motivation” can contain child nodes “Intrinsic Motivation” and “Extrinsic Motivation”.

Classifications

Classifications allow you to assign attributes to sources and participants. For example, for each respondent you can specify age, position, years of experience. This allows for subsequent comparative analysis by group.

Creating and Managing Codes in NVivo

Creating a New Code (Node)

1. Open the transcript in the Sources section
2. Highlight the desired text fragment
3. Right-click and choose Code Selection (or use a keyboard shortcut)
4. Select Code to New Node to create a new code or Code to Existing Node to apply an existing one
5. Enter the code name and, if desired, a description

Organizing Codes

• Drag and drop nodes to create a hierarchical structure
• Use the Merge Nodes function to combine duplicate codes
• Add descriptions to each node to create a codebook
• Use color marking for visual distinction of categories

Useful Functions for Analysis

• Word Frequency Query — analyze word frequency in data to identify key themes
• Coding Stripes — visual representation of all codes applied to a text fragment
• Matrix Coding Query — cross-analysis of codes by participant attributes
• Coding Comparison — comparing coding by several researchers (for checking inter-coder reliability)

Practical Assignments

Assignment 1: Coding Text

Read the following interview fragment and assign codes to it, using different coding types:

“When I first joined the company, everything seemed so complicated. Procedures, rules, unfamiliar people. But my mentor really helped—he literally ‘held my hand’ during the first three months. Now I have become a mentor myself and I understand how important that is for newcomers.”

Solution:

• Descriptive code: NEW EMPLOYEE ADAPTATION
• In vivo code: “HELD MY HAND”
• Process code: BECOMING A MENTOR
• Emotion code: FEELING OF CONFUSION (upon joining)
• Values code: VALUE OF MENTORSHIP

Assignment 2: Grouping Codes into Categories

Group the following codes into 2–3 categories and give each category a name: Unregulated working hours, Support from colleagues, Conflict with manager, Playing sports as relaxation, Teamwork, Urgent deadlines, Meditation, Toxic atmosphere, Socializing with friends.

Solution:

Category 1: STRESS FACTORS AT WORK

• Unregulated working hours
• Conflict with manager
• Urgent deadlines
• Toxic atmosphere

Category 2: SOCIAL SUPPORT

• Support from colleagues
• Teamwork
• Socializing with friends

Category 3: COPING STRATEGIES

• Playing sports as relaxation
• Meditation

Assignment 3: Working with NVivo

Describe step by step how you would organize a project in NVivo for a study on employee motivation if you have 15 interview transcripts from employees of three departments (sales, marketing, IT) of different levels (line staff and managers).

Solution:

1. Create a new project in NVivo with a descriptive title (for example, “Employee Motivation 2025”)
2. In the Sources section, create three folders by department: “Sales”, “Marketing”, “IT”. Import the transcripts into the appropriate folders.
3. Create Classifications for participants with attributes: Department (sales/marketing/IT), Level (line staff/manager), Years of experience, Age.
4. Assign classifications to each transcript.
5. Create an initial node (code) structure based on research questions: “Intrinsic Motivation”, “Extrinsic Motivation”, “Demotivating Factors”.
6. Conduct open coding of the first 3–4 interviews, creating new nodes as needed.
7. After initial coding, organize nodes hierarchically—group detailed codes under categories.
8. Use Matrix Coding Query to compare motivation patterns between departments and job levels.
9. Create a Memo for each key analytical decision, ensuring transparency of the research process.