Data Collection Methods, Tools and Research Instruments
Data collection converts research concepts into observations. The chosen method and instrument must match each objective, variable, participant group, setting and planned analysis while minimising burden and measurement error.
Learning objectives
- Compare major quantitative and qualitative collection methods.
- Develop questionnaires, interview guides and checklists.
- Operationalise variables and choose appropriate scales.
- Plan training, translation, field procedures and secure data handling.
Methods and tools
| Method | Instrument | Strength | Limitation |
|---|---|---|---|
| Survey | Questionnaire | Standardised data from many participants | Limited depth and response bias |
| Interview | Structured/semi-structured guide | Clarification and detailed experience | Interviewer influence and time |
| Focus group | Discussion guide | Norms and interacting viewpoints | Confidentiality and dominant voices |
| Observation | Checklist or field notes | Records actual practice and context | Observer effect |
| Record review | Extraction form | Efficient use of existing data | Missing and inconsistent documentation |
| Clinical measurement | Calibrated device/protocol | Objective physiological data | Calibration and technique error |
Questionnaire development
- Map every item to an objective and variable.
- Use validated items where appropriate and permitted.
- Write one clear idea per question.
- Choose response options that are exhaustive and mutually exclusive.
- Use logical ordering and skip instructions.
- Avoid leading, loaded, double-barrelled and ambiguous wording.
- Translate and back-translate when needed.
- Pretest, revise and document changes.
Worked item critique
Weak
“Are nurses friendly and competent?” combines two concepts and invites socially desirable agreement.
Improved
Use separate behaviour-based items: “Did the nurse explain your medicines clearly?” and “Were you treated respectfully?” with defined response scales.
Measurement scales
- Nominal: unordered categories such as blood group.
- Ordinal: ordered categories such as mild, moderate and severe.
- Interval: equal units without a true zero, such as temperature in °C.
- Ratio: equal units with a meaningful zero, such as age or weight.
Clinical measurement scenario
A blood-pressure protocol should specify device validation, cuff size, calibration, patient rest, posture, arm position, number and spacing of readings, recording rules and staff competency. “Measure blood pressure” alone is not reproducible.
Qualitative instruments
Use open-ended questions and neutral probes. Begin with broad, comfortable topics, progress to sensitive areas and close with an opportunity to add information. Record context, non-verbal observations and reflexive notes. Obtain specific permission for audio recording.
Fieldwork controls
- Standard operating procedures and interviewer training.
- Role-play, competency checks and pilot field practice.
- Unique IDs and daily completeness review.
- Range, consistency and skip checks in electronic forms.
- Calibration logs and duplicate measurements where appropriate.
- Encrypted storage, restricted access and separate identifier files.
Common errors
- Collecting interesting data unrelated to objectives.
- Using an unvalidated score without explaining adaptation.
- Offering overlapping response categories.
- Changing wording between participants.
- Recording identifiable information unnecessarily.
- Ignoring language, literacy, privacy and participant fatigue.
Review questions
- Choose tools for medicine adherence, provider practice and patient experience.
- Rewrite three leading or double-barrelled questions.
- Develop a reproducible blood-pressure measurement protocol.
- Compare interviews and focus groups for a sensitive topic.
References
- WHO. Health Research Methodology.
- UAHEB. Research Proposal and Report Writing Guidelines.
Educational content only; approved protocols and data-protection requirements must guide real studies.
