Data collection should produce the kind of evidence needed to answer the dissertation research question. The method should follow from the study design and evidence need rather than from convenience.
Start with the evidence the question requires
Ask what must be observed, measured, described or interpreted before choosing a technique.
The evidence need can usually be made more precise by asking what the final finding must contain. A study exploring experience may need detailed accounts and examples. A study measuring prevalence needs observations collected consistently across a defined group. A comparative question needs evidence that can be compared on the same dimensions, while a process question may require information about sequence, decisions and context.
This step prevents a common design problem: collecting information that is interesting but cannot answer the actual research question.
Choose the method after the design and sample
Qualitative methods, quantitative methods, interviews, questionnaires, focus groups, observations, documents and existing datasets solve different research problems. If selection is unclear, use sampling methods help.
Interviews are useful when participants need to explain experiences, reasoning or processes in their own words. Questionnaires can collect more standardised responses where the same variables or questions must be applied across respondents. Observation is useful when behaviour, interaction or process is more important than retrospective explanation. Documents and secondary datasets may be preferable when the required evidence already exists and can be accessed legitimately.
The choice can also involve trade-offs. A questionnaire may reach more participants but provide less explanatory depth. Interviews can produce rich evidence but require recruitment, recording, transcription and substantial analysis. The dissertation should explain why the selected trade-off is appropriate for its particular question.
Treat the instrument as part of the method
An interview guide, questionnaire, observation schedule or extraction form influences the evidence produced. Questions and measures should connect to the research objectives.
An instrument operationalises the data-collection plan. Interview prompts determine which experiences participants are invited to discuss. Questionnaire items determine how constructs are measured. Observation schedules determine what the researcher notices and records. Extraction forms determine which information is taken from documents or datasets.
Where appropriate, pilot testing can reveal ambiguous wording, missing response options, unrealistic interview length or recording problems before full data collection begins. Any pilot should follow the programme’s methodological and ethical requirements, especially when participant data are involved.
Describe procedure clearly
Explain access, timing, setting, recording and the steps that affect the evidence. Detail should clarify the research process rather than become a list of trivial actions.
A reader should be able to understand how evidence moved from access to usable research material. Relevant details can include how participants or records were approached, where collection occurred, how long it lasted, what instructions were given, whether recordings were made, and how completed responses or files were stored.
The aim is reproducible clarity, not administrative detail. Include a step when it can affect the quality, interpretation or ethics of the evidence.
Plan recording and data quality
Consider transcription, missing responses, inconsistent entries, duplicate records and other design-relevant quality issues.
Quality problems differ by evidence type. Interview research may need consistent transcription and clear speaker identification. Questionnaire data may contain missing responses, invalid values or duplicate submissions. Observational records can be affected by inconsistent recording, while secondary datasets may contain changing definitions or coding systems.
Planning these issues before collection makes later analysis more defensible because decisions about exclusions, corrections or missing evidence are not invented after the results are known.
Check feasibility and access
A method can be academically suitable but impossible within available time, permissions, recruitment access or data availability.
Feasibility should be tested before the method is finalised. Confirm that the required participants, organisations, documents or datasets can actually be accessed; that the collection can be completed within the available period; and that the student has the tools and skills needed to produce usable evidence.
If access is uncertain, the methodology should contain a realistic alternative rather than relying on evidence that may never become available.
Build ethics into collection
Consent, confidentiality, privacy, sensitive questions and data handling can affect how evidence may be collected. Use dissertation ethics help for deeper review.
Connect collection to analysis
The material collected must be suitable for the intended analysis. Continue to dissertation data analysis help for that stage. For the way these decisions fit together in the chapter, return to Methodology Chapter Help.