Software and Tools for Qualitative Analysis (Including the Role of SPSS)
❦
What Computer-Assisted Qualitative Data Analysis Software (CAQDAS) Does
CAQDAS packages store and organize data, code text, audio, video, and images, retrieve coded segments, query the data, link memos, and visualize relationships (Silver & Lewins, 2014). They do not analyze meaning; the researcher does (Bazeley & Jackson, 2013). Use them when you have many documents, long projects, teams, or a need for an audit trail.
Main Packages
| Software | Notes (as described in 2026 comparisons) |
|---|---|
| NVivo (Lumivero) | Deep, researcher-led coding, cases, queries, framework matrices, visualizations, and mixed-methods and survey integration; an AI Assistant offers summaries, coded-content summaries, and code suggestions (Lumivero, 2026; CASRAI, 2026) |
| ATLAS.ti (Berlin) | Flexible coding, networks and visual mapping, memos; extensive AI-assisted coding features, including coding directed by research questions; strong support for right-to-left languages in some comparisons (ATLAS.ti, n.d.; CASRAI, 2026) |
| MAXQDA (VERBI, Berlin) | Intuitive interface, integrated mixed-methods tools, visual tools, and an AI Assist feature for summarization and coding support (CASRAI, 2026) |
| Dedoose | Web-based; mixed-methods friendly; subscription |
| Taguette | Free and open source; text documents only; no hierarchical codes or visualizations (Skimle, 2026) |
| QualCoder | Free and open source (MIT licence) for text, audio, video, and images (Wikipedia, "CAQDAS," n.d.) |
| Other tools | QDA Miner, Quirkos, and R or Python packages for text analysis |
Many programs can exchange projects through the REFI-QDA (QDA-XML) standard (Wikipedia, "CAQDAS," n.d.). Prices and features change; student licences are often available through universities.
A Typical Workflow in Dedicated Software
| Step | NVivo | ATLAS.ti | MAXQDA |
|---|---|---|---|
| 1. Create a project | New project | New project | New project |
| 2. Import data | Import transcripts, PDFs, audio, video, survey data | Add documents | Import documents |
| 3. Record attributes | Case classifications with attributes | Document groups and attributes | Document variables |
| 4. Code | Create nodes and code by selecting text | Create codes and code quotations | Create a code system and code segments |
| 5. Write memos | Memos linked to nodes and sources | Memos | Memos |
| 6. Query and compare | Text search, coding queries, matrix coding queries, framework matrices | Code co-occurrence, networks | Code relations, visual tools |
| 7. Visualize | Charts, maps, word clouds | Network views | MAXMaps, code matrices |
| 8. Export | Codebook, reports, tables | Reports, exports | Reports, exports |
Always keep the codebook and a log of changes, and back up the project file regularly.
Free and Low-Cost Options
Use Word, Excel, or Google Sheets (columns for segment, code, theme; filters), or Taguette and QualCoder if your university does not provide a licence. These are sufficient for many student projects.
The Role of SPSS and Statistical Software
SPSS is a statistical package and is not designed for qualitative coding. Yet it has legitimate supporting roles:
- Describing the sample: frequencies and cross-tabulations of participant characteristics (in a participant table).
- Counting code frequencies after coding (in "small q" or content analysis designs) and testing associations, with caution.
- Analyzing open-ended survey responses after they have been coded into categories.
- Mixed-methods studies: the quantitative strand.
- IBM has offered text-analytics add-ons for SPSS in some versions; availability varies, so check your licence. These tools do not replace interpretive analysis.
If you are asked to use SPSS in a qualitative thesis, explain its limited, descriptive role in your methodology. In Big Q designs, do not convert themes into statistics to appear more "scientific".
Text Mining and Other Tools
R (for example the tidytext package) and Python can produce word frequencies, word clouds, sentiment analysis, and topic models from large text collections. These exploratory outputs can complement but not replace close reading.
Inter-Coder Agreement in Software
Programs can calculate Cohen's kappa (Cohen, 1960) or Krippendorff's alpha between coders. Landis and Koch (1977) give benchmarks (about 0.61 to 0.80 substantial; above 0.80 almost perfect). Such statistics suit codebook and content-analysis approaches; reflexive approaches instead use collaborative reflection and discussion rather than agreement statistics (O'Connor & Joffe, 2020; McDonald et al., 2019; Braun & Clarke, 2021).
Artificial Intelligence in Qualitative Analysis
Current QDA packages include AI features for transcription, summarization, and coding suggestions, and general-purpose generative AI tools can summarize or "code" text (Lumivero, 2026; CASRAI, 2026). The scholarly community is divided:
- Some researchers and guidance see AI as a supporting tool for transcription, summarizing, or an initial coding pass, provided the researcher reviews everything, and the use is transparent (Skimle, 2026).
- Others argue that GenAI is incompatible with reflexive qualitative research, because analysis is an interpretive, human, subjective practice, and they reject its use for that purpose; Jowsey et al. (2025), including the authors of reflexive thematic analysis, take this position. Concerns also include bias, privacy, hallucination, and lack of reflexivity.
Practical rules for students
1. Follow your university's AI policy and your supervisor's instructions, and obtain approval before using AI with data (Thesify, 2026).
2. Never upload identifiable or confidential data to public AI tools; check data-protection terms.
3. Retain interpretive control: the researcher reads the data and decides on codes and themes.
4. Verify any AI output against the original data.
5. Disclose the tool, version, purpose, and how you checked the output, in the methodology.
6. Match your approach to your methodology: in reflexive and interpretive designs, justify carefully, or avoid AI-generated analysis.
CHAPTER 16