Utilizing Software for Qualitative Data Analysis and Presentation

Qualitative research generates mountains of data – interview transcripts, field notes, open-ended survey responses, audio recordings, and more. Making sense of all that material manually is not just time-consuming; it is genuinely difficult to do rigorously and consistently. This is precisely why Computer-Assisted Qualitative Data Analysis Software (CAQDAS) has become an increasingly important part of the qualitative researcher’s toolkit. These programs do not replace the researcher’s thinking, but they make the heavy lifting of organizing, coding, and presenting qualitative data far more manageable.

What is CAQDAS? 🔗

Computer-Assisted Qualitative Data Analysis Software (CAQDAS) refers to a category of tools designed to help researchers organize, code, and analyze non-numerical data. Unlike statistical software packages used in quantitative research, CAQDAS does not perform the analysis for you. As a key insight from research published in the Malawi Medical Journal puts it, the primary function of CAQDAS is not to analyze data but to aid the analysis process – the researcher must always remain in control of interpretation.

CAQDAS is used across disciplines including psychology, sociology, marketing research, ethnography, and public health. At its core, these tools support tasks such as transcription analysis, coding and text interpretation, content analysis, discourse analysis, and grounded theory methodology. The CAQDAS Networking Project at the University of Surrey, formally established in 1994, has been instrumental in promoting the responsible and informed use of these tools in research communities worldwide.

A brief history: from NUD*IST to modern CAQDAS 🔗

The story of CAQDAS begins in the 1980s. The dawn of CAQDAS was marked by the development of NUD*IST – Non-numerical Unstructured Data Indexing, Searching, and Theorizing – a program specifically designed for qualitative data management. It was one of the first comprehensive tools that allowed researchers to move beyond handwritten index cards and paper-based coding systems.

NUD*IST introduced systematic coding and hierarchical organization of qualitative data. Researchers could build a tree-like structure of themes and sub-themes, then search and retrieve segments of data linked to any node in the tree. This was a significant leap forward. If you were studying workplace discrimination, for example, you could create a top-level node like “Discrimination Experiences,” with nested sub-nodes for “Race,” “Gender,” and “Age,” and then link every relevant passage from your interviews to the appropriate node. Searching the entire dataset for all passages coded under “Gender” would take seconds, not hours.

NUD*IST eventually evolved into NVivo, which its co-developer Tom Richards described as an improved and expanded version that introduced character-based coding, rich text capabilities, and multimedia functions. NVivo also added in-built facilities allowing researchers from different geographical locations to work on the same data files simultaneously through a network – a feature that has proven especially valuable in large collaborative research projects.

ATLAS.ti: a different approach to the same goal 🔗

While NUD*IST/NVivo organizes data in a hierarchical, nested structure, ATLAS.ti takes a notably different approach. ATLAS.ti’s organization is flat – all documents and codes exist in the same space, but they can be grouped together flexibly through user-defined groups and code groups. This makes it easier to assign a single document or code to multiple categories at once, which can better reflect the messy, overlapping nature of real-world qualitative data.

One of ATLAS.ti’s most recognized features is its ability to create visual network diagrams that show relationships between codes, quotations, and memos. A researcher studying the lived experiences of first-generation college students, for instance, could generate a network map showing how the theme “financial stress” connects to “family obligations,” “academic performance,” and “identity conflict.” These visual representations can reveal structural patterns in the data that are easy to miss when working through text alone.

Modern versions of ATLAS.ti also include AI-assisted coding capabilities that can suggest codes, identify sentiment, and highlight named entities such as people, places, and organizations – while keeping the researcher in full control of what is accepted or refined. This positions it as a strong option for large-scale or mixed-methods projects that require both analytical depth and flexibility.

Core functions shared by CAQDAS tools 🔗

Whether a researcher chooses NVivo, ATLAS.ti, MAXQDA, Dedoose, or another platform, most CAQDAS tools share a set of fundamental capabilities that transform how qualitative research is conducted.

Data management and organization 🔗

Qualitative analysis tools serve as workbenches where the general tasks of qualitative work can be done – from organizing and managing qualitative materials including text, audio, video, images, and social media data, to marking up and tagging those materials systematically. Researchers can import raw data in various formats – from plain text and PDFs to audio and video files – and keep everything in one organized project environment.

Coding and theme identification 🔗

Coding is the most fundamental strategy in qualitative research, and all CAQDAS tools center on this function. This involves the application of a maintained set of terms and short phrases linked to segments of text or audio/video that can then be queried and gathered for comparative analysis. Researchers can apply codes manually, or in more advanced tools, use auto-coding features that flag words or phrases when they appear across documents. This speeds up initial pass-throughs considerably without sacrificing researcher judgment in the final interpretation.

Querying and pattern analysis 🔗

Once data is coded, CAQDAS tools allow researchers to run queries across the entire dataset. Want to know every instance where a participant mentioned “job insecurity” in the context of “mental health”? A Boolean query across a coded dataset can retrieve all such passages instantly. Advanced Boolean search options and flexible interlinking of segments, codes, and annotations allow for deep pattern discovery that would be nearly impossible to replicate by hand across a large dataset.

Visualization and presentation of findings 🔗

Modern CAQDAS tools offer a range of visualization options that make research findings easier to present and understand. These include mind maps, word clouds, and charts that provide intuitive overviews of dominant themes and their relationships. Network diagrams are particularly useful for showing how different concepts relate to one another, while matrix displays allow side-by-side comparisons of how different participant groups responded to the same topics. For longitudinal studies, some platforms can even generate timelines showing how themes shifted over time.

NVivo vs. ATLAS.ti: choosing the right tool 🔗

Both NVivo and ATLAS.ti are widely used in academic, government, and applied research environments, and both handle diverse data types – text, PDFs, audio, video, and survey data – within a single analysis environment. The choice between them often comes down to project needs and researcher preference.

NVivo largely thinks about project organization in a nested, hierarchical fashion: files can be organized in folders with sub-folders, and codes have a parent-child structure. This makes it particularly strong for researchers who prefer a structured, top-down analytical framework. NVivo is geared towards well-funded academics and research professionals handling exceptionally large datasets, though some users find it less collaboration-friendly due to its desktop-first design.

ATLAS.ti, by contrast, suits researchers who prefer a more flexible, iterative approach to coding – particularly those working with visual or multimedia data and those who need strong network visualization capabilities. ATLAS.ti is a flexible and methodologically robust CAQDAS platform used across academic, government, and applied research environments, and it supports both qualitative and mixed-methods designs.

For students or researchers new to CAQDAS, more accessible and affordable tools such as Dedoose, Quirkos, or the free open-source option Taguette offer entry points into qualitative data analysis without the steep learning curve of the more advanced platforms.

The researcher’s role in software-assisted analysis 🔗

A critical point that deserves emphasis is this: CAQDAS does not analyze data – researchers do. The software manages, organizes, and retrieves; the intellectual work of interpretation remains entirely with the researcher. Concerns about CAQDAS include the risk of increasingly deterministic and rigid processes, a privileging of coding and retrieval over deeper meaning, and increased pressure to focus on volume and breadth rather than depth. These are legitimate cautions, not reasons to avoid the tools, but reminders that the software should serve the research – not the other way around.

The transparency CAQDAS provides is one of its most undervalued benefits. Using CAQDAS has been recognized as aiding the researcher in producing an accurate and transparent picture of the data while also providing an audit of the data analysis process as a whole – something that has often been missing in accounts of qualitative research. This audit trail allows other researchers to scrutinize the analytical process, strengthening the credibility and reproducibility of qualitative findings.

Additionally, NVivo supports code-based inquiry, searching, and theorizing combined with the ability to annotate and edit documents – functions that help researchers document their interpretive decisions as they go, building a transparent record of how conclusions were reached.

Practical steps for using CAQDAS in a research project 🔗

Using CAQDAS effectively mirrors the traditional qualitative research process but with added efficiency. The typical workflow moves through several stages. First, raw data is imported – transcripts, recordings, or field notes – and the researcher reads through everything to develop an initial sense of the material. Then comes systematic coding, where meaningful segments are labeled. This may be done line-by-line for fine-grained analysis, or in larger thematic chunks. As coding progresses, the researcher runs queries, builds memos, and refines the coding framework. Finally, visualizations are generated to support the write-up and presentation of findings.

In 2019, the Rotterdam Exchange Format Initiative (REFI) launched the QDA-XML open exchange standard, which allows researchers to move coded qualitative data from one software package to another. This development significantly reduces the lock-in risk of committing to a single platform and encourages greater collaboration and interoperability across the qualitative research community.

Ultimately, the value of CAQDAS is not in which specific program you choose – it is in developing the discipline to use any of these tools systematically, transparently, and in genuine service of answering your research question.

What do you think? As qualitative research increasingly intersects with digital and AI-assisted tools, where should the boundary lie between software assistance and researcher interpretation? And do you think the move toward visual, software-generated outputs changes how qualitative findings are perceived in academic and policy settings?

References

1. https://en.wikipedia.org/wiki/Computer-assisted_qualitative_data_analysis_software
2. https://pmc.ncbi.nlm.nih.gov/articles/PMC4478399/
3. https://www.surrey.ac.uk/computer-assisted-qualitative-data-analysis
4. https://latisresearch.umn.edu/qualitative-analysis-software
5. https://lumivero.com/resources/blog/top-caqdas-tools-for-qualitative-research/
6. https://guides.library.jhu.edu/QDAS
7. https://delvetool.com/blog/atlasti-vs-nvivo-vs-delve
8. https://www.tandfonline.com/doi/full/10.1080/13645579.2020.1803528