How many interviews? A systematic review and mixed-methods synthesis to guide sample size in qualitative research
T Hampton, L Nkhalamba, S Salimu, M Bhutta, D NyirendaINTRODUCTION
Guidance for qualitative research involving participant interviews remains unclear around a key logistical question: how many interviews are required? Sample size may depend on several variables but researchers need to propose a suggested number of interviews at multiple stages before research is conducted. We examined the literature for evidence to support decisions and recommendations for sample size in health-related qualitative interview research.
METHODS
A systematic review was conducted of eight electronic databases, with no language or time restrictions, for evidence or guidance on sample sizes for qualitative interviews. A mixed-methods meta-narrative approach was adopted for evidence synthesis.
RESULTS
A total of 58 studies were included in the final analysis. There was a lack of consensus on recommended sample size, reflecting the wide number of belief systems and methodologies that use qualitative interviews. The mean and median lowest recommended figures from included studies were 10 and 9 interviews respectively. Depending on question and epistemology, evidence for sample sizes from 1 to 298 was found. A saturation-based justification was used by many studies. Our analysis determined that readers seeking to use a rule of thumb could base this either on precedent-justified figures or frequentist and empirical ‘saturation’-based numbers of interviews, described in the Discussion section.
CONCLUSIONS
Our review of justifications for number of interviews for qualitative research found heterogeneity based on epistemology and standpoint. This review should guide surgeons and other healthcare workers on initial calculations for interview sample sizes. A preponderance of English language and empirical justifications could lead to epistemic injustice that further research in the field should seek to address.