Qualitative vs Quantitative vs Mixed Methods: How to Decide
The most fundamental design decision in doctoral research — explained through the questions each approach can actually answer.
Published 15 January 2025 · Updated 15 January 2025 · 9 min read · By the MaquishTech Venture® research consulting team
What each approach can — and cannot — answer
Quantitative research answers questions of magnitude and relationship: how much, how often, is there a difference, is there a correlation, does X predict Y. It requires measurable variables and produces numbers that can be statistically tested.
Qualitative research answers questions of meaning and process: how do people experience this, why do they behave this way, what does this phenomenon look like from inside. It produces themes, narratives and explanations rather than numbers.
Mixed methods combine both — but only when the research question genuinely requires both kinds of evidence. "I did a survey AND an interview" is not mixed methods unless the two strands are deliberately integrated.
Decision criteria
- Question asks "how much / is there a difference" → quantitative
- Question asks "how / why / what is the experience" → qualitative
- Question has both strands AND they must inform each other → mixed methods
- Existing theory is strong → quantitative hypothesis testing fits
- Little is known about the phenomenon → qualitative exploration first
- You need breadth (300 responses) → quantitative; depth (15 interviews) → qualitative
Common mixed-methods designs
The three canonical designs: explanatory sequential (survey first, interviews to explain the numbers), exploratory sequential (interviews first, then a survey to test themes at scale), and convergent (both collected in parallel, merged in interpretation).
Choose sequential designs when one strand must inform the design of the other. Choose convergent when time is short — but budget analysis time for genuine integration, which is where most mixed-methods studies actually fail.
Mistakes reviewers catch
- Calling a survey with two open-text questions "mixed methods"
- Collecting qualitative data but analysing it with word counts instead of coding
- Using a convenience sample but writing as if it were representative
- Reporting statistical significance without effect size or confidence intervals
- Treating qualitative validity terms (credibility, transferability) as interchangeable with quantitative ones (reliability, validity)
Frequently asked questions
Working on your own research?
We provide ethical research coaching: methodology guidance, data-skills training, literature-search strategy, proof-reading and review of your own work. You remain the sole author and you do the analysis. We do not complete, or offer to complete, any part of assessed work. See Ethical PhD Research Mentoring.
