Qualitative research produces a lot of material: interview transcripts, focus group recordings, field notes, documents, open-ended survey responses and images. Keeping track of all of it in Word files and spreadsheets quickly becomes difficult. Computer-assisted qualitative data analysis software (often called CAQDAS) such as NVivo, ATLAS.ti and MAXQDA helps you store, organise, code, search and retrieve your data in one place.
This post explains what qualitative data analysis software can and cannot do, how to set up and organise a project in NVivo step by step, how other software options compare, and good practices that will make your analysis easier to manage and to defend.
What qualitative software does, and what it does not do
A common misunderstanding is that the software analyses the data for you. It does not interpret meaning or build themes on its own. You still do the thinking. What the software does well is:
- Store all your data sources in a single project file.
- Organise data by participant, site, group or any other attribute.
- Code segments of text, audio, video or images.
- Retrieve everything coded at a particular code across all sources.
- Query the data, for example to compare how different groups talk about a theme.
- Record your thinking through memos and annotations.
- Visualise relationships between codes, cases and themes.
- Document your analytic process, which supports an audit trail.
Setting up an NVivo project step by step
Menu names can change between versions, so check the help documentation for the version you are using. The general workflow is similar across versions.
1. Prepare your data before importing
- Use a consistent file naming system, such as P01_Interview_Manager, P02_Interview_Employee.
- Format transcripts consistently, with speaker names and, if useful, headings for each interview question.
- Remove or replace identifying information to protect participant confidentiality.
- Keep a master copy of raw files outside the software.
2. Create a new project and import your files
Create a project with a clear name and save it in a secure, backed-up location. Import your files, such as transcripts, PDFs, audio or video files, images and spreadsheets of survey responses. Use folders to separate data types, for example Interviews, Focus Groups and Documents.
3. Create cases and attributes
Cases represent your units of analysis, such as individual participants or organisations. Assign each source to a case and add attributes (also called classifications), such as gender, role, years of experience or organisation type. Attributes are very useful later, because they let you compare how different groups talk about the same codes.
4. Build your coding structure
Codes (called nodes in older versions of NVivo) are containers for segments of data about a particular idea. You can create codes in advance from your research questions or theory (deductive coding), create them as you read the data (inductive coding), or combine both. Organise codes hierarchically, with broader parent codes and more specific child codes.
5. Code your data
Read each source carefully, select a segment of text or media, and assign it to one or more codes. Keep code names short and clear, and write a definition for each code in its description field. This becomes your codebook.
6. Write memos as you go
Use memos to record your thoughts, emerging interpretations, questions, and decisions about merging or splitting codes. Link memos to relevant sources or codes. Memos are the bridge between coding and writing your findings.
7. Review and refine codes
Open each code to read all the segments coded there. Ask whether the segments really belong together, whether the code should be split or merged, and whether the definition needs updating. This review is where codes start to develop into categories and themes.
8. Run queries
Queries help you explore patterns. Useful examples in NVivo include:
- Word frequency query: shows the most frequently used words, which can help you get familiar with the data.
- Text search query: finds every occurrence of a word or phrase.
- Coding query: retrieves content coded at combinations of codes.
- Matrix coding query: cross-tabulates codes against cases or attributes, for example how often managers and employees mention a particular theme.
- Coding comparison query: compares coding by different researchers, useful if you are calculating inter-rater reliability.
9. Visualise your analysis
Use mind maps, concept maps, hierarchy charts or project maps to explore and present relationships between codes and themes. These visuals can also support your discussion with supervisors.
10. Export and report
Export your codebook, coded extracts and query results to use in your thesis. Select strong, representative quotations for each theme and keep a link between each quotation and its source.
Other software options
- ATLAS.ti: a widely used alternative with strong network visualisation features and support for many data types.
- MAXQDA: popular for qualitative and mixed-methods research, with tools for integrating qualitative and quantitative data.
- Dedoose: a web-based tool designed for collaborative and mixed-methods projects.
- Taguette and QualCoder: free, open-source options that cover the core functions of importing, coding and retrieving data.
- Word processors and spreadsheets: for small projects, careful use of comments, highlighting and tables can work well.
Most commercial packages offer student pricing or free trials, and many universities provide institutional licences. Check what your institution supports, since training and technical help may be available.
How to choose the right tool
- Size of your data: a few interviews may not need specialised software; dozens of interviews or mixed media usually do.
- Type of data: check support for audio, video, images, PDFs or social media data if you need them.
- Collaboration: if several researchers will code together, check team and merging features.
- Operating system and cost: confirm compatibility and licence availability.
- Supervisor and institutional support: using the same tool as your supervisor or research group makes sharing and feedback easier.
Good practices for organising qualitative data
- Back up regularly to a secure, approved location, and keep version-dated copies of your project.
- Protect confidentiality by anonymising data and following your ethics approval and data management plan.
- Keep a codebook with definitions, examples and dates of changes.
- Avoid creating too many codes too early; review and consolidate regularly.
- Write memos frequently to record your reasoning.
- Do not let counts replace interpretation. How often a code appears is not the same as how important it is.
- Describe the software's role accurately in your methodology chapter: it supported data management and coding, while the analysis was conducted by the researcher.
Final thoughts
NVivo and similar software can make qualitative analysis more organised, transparent and efficient, especially in larger projects. Prepare your data carefully, set up cases and attributes, build a clear coding structure, write memos, and use queries and visualisations to explore patterns. Remember that the software is a tool for managing your analysis, not a substitute for your own careful reading and interpretation.