triangulation

# Triangulation in Qualitative Research Triangulation uses **multiple vantage points** to strengthen interpretation. It can deepen understanding—but it is not a mechanical “more is always better” rule, and it does not automatically produce a single “true” reality. ## Types of triangulation - **Data triangulation:** Multiple sources (interviews, observations, documents, artifacts). - **Investigator triangulation:** Multiple analysts code/discuss the same material. - **Theory triangulation:** Multiple theories/lenses inform interpretation (use carefully to avoid eclectic forcing). - **Methodological triangulation:** Multiple methods (e.g., interviews + ethnography; qualitative + quantitative in mixed methods). ## Denzin’s framework (classic) Denzin distinguished these modes as ways to overcome partiality of single methods/sources. Contemporary writers emphasize triangulation for **completeness and depth**, not only convergence. ## Purposes - **Confirmation:** Independent sources support the same inference (handle cautiously—agreement can reflect shared bias). - **Completeness:** Different methods capture different facets (e.g., policy text vs shop-floor practice). - **Complementarity:** Methods illuminate mechanisms vs prevalence, meanings vs distributions. ## How to conduct each type (practically) **Data triangulation:** Build an **evidence table** linking claim → supporting excerpts across sources → disconfirming evidence searched. **Investigator triangulation:** Use structured coding comparisons, reconciliation sessions, and documented rule changes for the codebook. **Methodological triangulation:** Pre-specify **integration points** (joint displays, threading narratives with descriptive statistics) so mixing is analytic, not decorative. ## Triangulation in grounded theory GT already “triangulates” through **constant comparison** across incidents. Adding sources should serve **theoretical sampling** aims: elaborate categories, conditions, and consequences. Avoid importing a fixed multi-method plan that steers analysis away from emergent core categories. ## Limitations and critiques Triangulation assumes a somewhat realist stance when framed as convergence. Constructivist researchers may prefer **crystallization** metaphors: multiple facets refract light differently. Also, triangulation increases **cost, time, and IRB complexity**—justify it analytically. ## Output format (example) For a finding F: | Source type | Evidence snippet ID | Supports? | Notes / tensions | |-------------|---------------------|-----------|------------------| | Interview | P12: 00:18:22 | Yes | Describes mechanism | | Document | Policy §4.2 | Partial | Official vs practiced gap | | Observation | Site visit 3 | No | Suggests alternative routine | ## Checklist - [ ] Triangulation type matches research question and paradigm. - [ ] Integration strategy defined (not just data pile-on). - [ ] Disconfirming evidence actively sought across sources. - [ ] GT: triangulation serves theoretical sampling, not premature closure. - [ ] Limitations of convergence logic acknowledged where constructivist. ## References (starting points) - Denzin, N. K. *The research act* (classic triangulation typology—read contemporary critiques too). - Flick, U. *An introduction to qualitative research* — triangulation in practice. - O’Cathain, A., Murphy, E., & Nicholl, J. (2010). Multidisciplinary, interdisciplinary, or dysfunctional? *Journal of Mixed Methods Research*.