Mixed Methods Research
Mixed methods research combines quantitative and qualitative approaches within a single study or programme of inquiry. Rather than treating the two paradigms as incompatible, mixed methods researchers argue that combining them yields a more complete understanding of the research problem than either approach alone.
Core Designs
Creswell & Plano Clark (2018) identify several primary designs:
| Design | Structure | When to use |
|---|---|---|
| Convergent (concurrent) | QUAN + QUAL collected simultaneously, then merged | When you want different lenses on the same phenomenon at the same time |
| Explanatory sequential | QUAN → QUAL | When quantitative results need qualitative explanation (e.g., why did the treatment group improve?) |
| Exploratory sequential | QUAL → QUAN | When qualitative findings inform instrument development or hypothesis generation |
| Embedded | One strand nested within the other | When a secondary dataset supplements the primary design (e.g., interviews within an experiment) |
Upper-case (QUAN/QUAL) indicates the dominant strand; lower-case indicates the supplementary strand.
In Applied Linguistics
Mixed methods research has grown rapidly in SLA and language teaching since the 2000s. Common applications include:
- Intervention studies — quasi-experimental pre/post-test data combined with learner interviews, stimulated recall, or classroom observation to explain why the treatment worked (or did not)
- Assessment research — statistical analysis of test scores combined with think-aloud protocols to investigate test-taking processes
- Teacher development — surveys establishing patterns across a population, followed by case studies of individual teachers
- Programme evaluation — achievement data plus stakeholder perspectives
Strengths
- Compensates for the weaknesses of each approach: quantitative breadth + qualitative depth
- Triangulation across methods strengthens claims
- Provides both what happened (quantitative) and why/how (qualitative)
- Increasingly valued and expected by journal editors and funding bodies
Challenges
- Requires competence in both paradigms — statistical analysis and qualitative coding
- Time-intensive: two data sets to collect, analyse, and integrate
- Integration is the key challenge — many studies collect both types of data but fail to genuinely merge findings
- Paradigmatic tensions — some argue that positivist and interpretivist assumptions are fundamentally incompatible (the "paradigm wars")
Quality Criteria
Teddlie & Tashakkori (2009) proposed "inference quality" and "inference transferability" as mixed methods equivalents of validity and generalisability. The quality of integration — how well the quantitative and qualitative strands inform each other — is the defining criterion.
Key References
- Creswell & Plano Clark (2018) — Designing and Conducting Mixed Methods Research (3rd ed.)
- Teddlie & Tashakkori (2009) — Foundations of Mixed Methods Research
- Hashemi & Babaii (2013) — mixed methods in applied linguistics
- Riazi & Candlin (2014) — mixed methods research in language teaching and testing