Triangulation

Triangulation is the use of multiple data sources, methods, researchers, or theoretical perspectives to study the same phenomenon. The logic is convergence: if different approaches yield consistent findings, confidence in the results increases. If they diverge, the discrepancy itself is analytically productive.

Denzin's Four Types

Norman Denzin (1978) identified four types of triangulation, which remain the standard typology:

Type	Description	SLA example
Data triangulation	Multiple data sources (different times, settings, or participants)	Collecting data from beginners, intermediates, and advanced learners
Investigator triangulation	Multiple researchers analyse the same data	Two coders independently coding Corrective Feedback episodes
Theoretical triangulation	Multiple theoretical frameworks applied to the same data	Interpreting classroom interaction through both sociocultural and cognitive lenses
Methodological triangulation	Multiple methods used in the same study	Combining test scores with interviews and Classroom Observation

Denzin further distinguished within-method triangulation (using multiple scales within a questionnaire) from between-method triangulation (combining quantitative and qualitative methods, essentially Mixed Methods Research).

Purpose

Triangulation serves different purposes depending on the research paradigm:

• In post-positivist research: convergent results increase confidence that findings are "true" (closer to traditional validity)
• In interpretivist research: multiple perspectives enrich understanding without claiming one objective truth
• In critical research: different data sources can reveal contradictions between what participants say and what they do

In Applied Linguistics

Triangulation is particularly valued in:

• Classroom-based Research: test scores alone cannot explain classroom dynamics; adding interviews and observation creates a fuller picture
• Teacher Cognition research: comparing what teachers say in interviews with what they do in observed lessons (Teacher Beliefs vs practice)
• Assessment validation: multiple evidence sources for Construct Validity
• Action Research: teachers triangulate data from student work, observations, and reflections

Cautions

• Triangulation does not guarantee validity: if all methods share the same bias, convergence is misleading
• Divergent findings are not necessarily a problem; they may reveal complexity
• The term is sometimes used loosely to mean "I used multiple methods" without genuine integration of findings

References

• Denzin (1978): The Research Act (four types of triangulation)
• Lincoln & Guba (1985): triangulation as a credibility strategy in Qualitative Research
• Creswell & Miller (2000): validity procedures in qualitative research including triangulation