Group evidence by a clear rule, then return to the original context before drawing a conclusion.
Qualitative evidence describes meanings and experiences. Read complete interview notes before coding them. A code is a label for a relevant idea, such as safety, cost or social contact. Write a short definition and apply it consistently; one passage may receive more than one code.
Compare themes across participants and places. Count mentions when useful, but state whether you count words, passages or people. A person repeating "hot" five times is not five residents reporting heat. A word cloud shows frequency according to a chosen counting rule, not the importance or meaning of an issue.
Keep contrary cases. Some residents may value a busy market for company while others find it stressful. Quote briefly and anonymously where consent permits, and explain what the words show in relation to the question. Do not treat one vivid comment as proof of a population-wide view.
Compare coding between investigators to identify ambiguous categories. Use observations and numerical patterns alongside interviews where relevant. Agreement can strengthen an interpretation, while disagreement can reveal a difference in timing, definition or experience.
Step by step
Define the codes
State which words or meanings qualify for each category.
Compare fairly
Use the same rule and retain examples that disagree with the main theme.
Interpret in context
Connect the theme to the question and the participant's circumstances.
Worked example: A transparent coding rule
Code "mobility barrier" when a participant describes physical difficulty moving along a route. Do not include every negative comment: high prices belong under cost unless the participant links them to travel access.
Watch out for this
The most frequent word must be the main problem.
Frequency depends on wording and counting rules. Importance also depends on severity, context and who is affected.
Check your understanding
One interview contradicts the dominant theme. What should you do?
- Delete it so the conclusion is clearer.
- Examine its context and explain the difference.
- Discard all other interviews.