Choosing Between Qualitative and Quantitative Methods

The method should follow the question, available data and ethical requirements. Quantitative approaches can estimate patterns with numerical measures; qualitative approaches can explore experiences, meanings and processes in depth. Neither is automatically more rigorous. A project can also combine approaches, but doing so creates extra work and requires a clear reason. Begin by stating what you want to know and what kind of evidence would make an answer credible.

Match the question to the evidence

A question about how many students use a service needs a suitable count or survey with a defensible sample. A question about how students experience that service may be better suited to interviews or observation. A question about why outcomes differ might require several forms of evidence. Avoid choosing a method because it sounds impressive; ask whether its data could actually answer your specific question.

Consider sampling and access

A large dataset can still be misleading if it excludes the population you care about. A small set of interviews can provide valuable detail while offering limited grounds for broad numerical estimates. State where participants or documents come from and what access you realistically have. If you cannot obtain permissions before the deadline, a secondary-data project may be more appropriate if the brief allows it.

Plan analysis before collecting

Decide how you would analyse the evidence before designing the instrument. For numbers, define measures, comparisons and missing-data decisions. For interviews, identify an initial approach to coding, comparison and documenting interpretations. A mixed-methods project needs to explain how the strands will inform one another, not simply place an interview quote beside a chart.

Address quality and ethics

Check validity and reliability of measures where relevant, and credibility and transparency of qualitative interpretation. Consider consent, privacy, secure storage and any institutional review requirements before approaching participants. Pilot a questionnaire or interview guide to reveal unclear wording. Do not promise anonymity if combinations of details could identify someone. Your supervisor’s rules take precedence over a general guide.

Explain the choice in the methods section

State what the question requires, why the selected approach fits and what it cannot establish. “Interviews were easier” is a logistical explanation, not an analytical justification by itself. If access constraints shaped the design, acknowledge them. Explain enough of the sampling, collection and analysis process for another reader to understand how your findings were produced.

Worked example

To study first-year students’ experiences of library induction, interviews could reveal which steps caused confusion, while a properly designed usage survey could estimate how common a reported problem is. If your project has only six volunteer interviews, report themes from those accounts; do not turn them into a percentage for the whole university. A small project can still be useful when its claims fit its method.

A method-choice decision table

Write the research question in the first column. In the next columns, list the kind of claim desired, evidence needed, feasible collection route and main limit. For “How common is this experience?”, a defensible sampling plan matters. For “How do participants describe the experience?”, depth and interpretation matter. If your available evidence is six volunteer interviews, consider whether to change a prevalence question into an exploratory one. If you have administrative counts but no accounts of decision-making, avoid claiming to know motives. This table helps a student avoid a common mistake: starting with a favourite technique and bending the question until it fits. Show the finished choice to the supervisor before recruitment, particularly when data involve people or sensitive records.

Checklist before you submit

State the question before choosing the method.
Check access, sampling and analysis feasibility.
Plan ethics and data protection before collection.
Describe limitations in the same terms as the claims.

Questions students ask

Can I use both approaches?

Yes when each answers a distinct part of the question and you can explain how the evidence will be integrated.

Does a larger sample guarantee better research?

No. Sampling, measurement and fit to the question also matter.