Interpreting the Data
❦
Analysis organizes the data; interpretation explains what they mean. It is the heart of a qualitative study and the stage at which examiners judge your insight.
From Description to Interpretation
| Level | Question | Example |
|---|---|---|
| Description | What did participants say or do? | "Many owners said they feared system failures." |
| Analysis | What patterns and relationships are in the data? | "Fear was linked to lack of immediate technical support." |
| Interpretation | What does this mean, and why does it matter? | "Adoption depends less on perceived usefulness than on the availability of trusted help, which suggests that support networks act as a precondition for trust." |
Principles of Interpretation
1. Stay close to the data, and show the evidence for each claim.
2. Answer the research questions explicitly.
3. Go beyond topics to meaning (themes, not just categories).
4. Consider alternative explanations, and say why you prefer yours.
5. Examine variation: who is different, and why?
6. Link to context: the setting, culture, and history.
7. Link to theory and literature: where do findings confirm, extend, or challenge them?
8. Acknowledge your influence (reflexivity).
9. Distinguish what participants said from your interpretation.
10. Avoid overclaiming: do not generalize to populations or imply causation without justification.
Semantic and Latent Interpretation
Semantic interpretation stays at the surface of what was said; latent interpretation examines underlying assumptions, ideologies, and meanings (Braun & Clarke, 2006). Both are legitimate, but be clear which you are doing, and ensure that it fits your paradigm.
Building Themes and Models
- A theme has a central organizing concept and a story. Test it by asking, "Can I state what this theme is about in a sentence?"
- A thematic map or table shows themes, sub-themes, and relationships.
- In grounded theory, a core category and a model explain the process (for example conditions, actions, consequences).
- In case studies, propositions or cross-case patterns summarize the explanation.
An example of a theme table
| Theme | Sub-themes | Brief description |
|---|---|---|
| From fear to confidence | Initial apprehension; learning by doing; growing trust | Owners moved from fear of failure to confidence as they gained experience |
| Dependence on trusted help | Family support; peer advice; provider support | Adoption relied on informal and formal helpers |
| Reasons for continuing | Customer demand; safer cash handling; record keeping | Practical benefits sustained use |
Using Quotations
- Select quotes that are vivid and illustrate the theme; avoid using quotes only as decoration.
- Introduce each quote and interpret it afterward.
- Edit lightly and mark any omissions with ellipses; do not change meaning.
- Label with a pseudonym or code (for example P4, grocery owner).
- Length: short quotes are integrated into the text; longer ones (in APA, 40 or more words) are set as block quotations.
- Balance: use quotes from several participants, not only the most articulate (Sandelowski, 1994; Corden & Sainsbury, 2006).
- Translation: indicate that quotes were translated, and how.
An Analytic Paragraph Formula
Assertion (the claim), evidence (quotes or data), interpretation (what the evidence means), and link (to the question, literature, or theory).
Example: "Participants described a transition from fear to confidence. P4 explained, 'At first I was afraid. If the machine failed, who would help me?' Yet the same participant later said that customers 'prefer it'. This shift suggests that confidence was built through experience and social proof rather than through a prior belief in usefulness, which extends technology adoption models that emphasize perceived usefulness (Davis, 1989)."
Visual Displays
Matrices, process diagrams, timelines, and networks can clarify your analysis (Miles et al., 2020). Make sure each is explained in the text.
Counting in Qualitative Research
Reporting that "most", "several", or "a few" participants said something is common, but define these terms (for example, "most" meaning more than half). Avoid percentages with small samples. Frequency does not equal importance.
Common Mistakes
- Reporting topics instead of themes.
- Using quotes without interpretation.
- Over-quantifying.
- Ignoring contradictory data.
- Forcing data into a pre-existing theory.
- Confusing your assumptions with participants' views.
CHAPTER 18