Annie Fisher & Sharon James 2008
Considering triangulation as a means of strengthening data in interpretive research
1. Triangulation: Qualitative research is inherently multimethod in focus. Flick (2002:226)
A range of methodological literature (for example, Denscombe, 2003; Flick, 2002; Silverman, 2000, 2006; Wellington, 2000) suggests that conducting interpretive research is inextricably bound up with collecting multiple versions of ‘reality’ or ‘truth’.
Without this series of lenses, it is suggested that the resulting uni-dimensional view of the situation is inadequate in providing the researcher with the means to ‘interpret’ what is perceived.
2. A Basic Definition: “An attempt to map out, or explain more fully, the richness and complexity of human behavior by studying it from more than one standpoint“ Cohen, Manion and Morrison (2000:112)
Silverman’s later definition provides the analogy with navigation:
"Comparing different kinds of data (e.g. quantitative and qualitative) and different methods (e.g. observation and interviews) to see whether they corroborate one another. This form of comparison, called triangulation, derives from navigation, where different bearings give the correct position of an object". Silverman (1993:156)
In more recent writing on interpreting qualitative data, Silverman (2006) suggests that whatever the researcher’s theoretical stance, the issue of validity is central. In interpretive research, this combination of “a variety of theories, methods, observers, and empirical materials to produce a more accurate, comprehensive and objective representation of the object of study” (p.201) has often been considered to be the most effective way of ensuring validity; it was assumed that if a variety of methods allowed the same conclusions to be drawn, validity had been established.
3. The interpretive researcher: often seeks to address validity and reliability through the use of triangulation. This can lead to the misguided search for a single truth. Mason (1995)
According to Eisner (1998:55), in any research “the need for replicability is a need for reliability” as the “need for relevance is the need for validity”; he describes the use of coherence, corroboration and consensus in the evaluation of qualitative research as “structural corroboration”: the use of multiple data sources so that conclusions can be “structurally corroborated”. This is clearly similar to triangulation.
Mason (1995) suggests methodological triangulation as a way of corroborating the data in qualitative case studies. This implies that we may obtain a ‘true’ understanding of the situation by seeking the intersection of, for example, interview and observational data though a process of ‘overlaying’. This has clear links to the vexed question of ‘truth; however, as Silverman (1993) cautions, if we conceive reality as socially constructed, and context-dependent, then no one ‘phenomenon’ can be applied to all cases to provide a definitive and objective explanation.
Silverman (2006) and others (for example: Cresswell, 2007; Fielding and Fielding, 1986; Mason, 1995, 2000; Flick, 2002), however, posit that, contrary to Denzin’s(1970) original theorizing on method triangulation (see slide 4) as a means of finding ‘truth’, this is not adequate in social research In a quest for reliability, Silverman (1993) cautions that if we conceive reality as socially constructed, and context-dependent, then no one ‘phenomenon’ can be applied to all cases. We also need to be aware of the naiveté inherent in a search for ultimate ‘truth’ through triangulation.
4. Denzin’s Typology: Data triangulation – time, space, person; investigator triangulation; theory triangulation; methodological triangulation. (1970, In Wellington,2000)
As an early advocate of triangulation, Denzin’s (1970) original typology is much discussed (Wellington, 2000; Denscombe, 2007). To overcome partial views of any ‘case’, Denzin suggests using data triangulation, subdivided into (1) time: considering cross-sectional and longitudinal studies; (2) space: engaging in comparative study of populations, countries, and (3) person: at the level of the individual, the interaction between groups and the collection level. He goes on to suggest investigator triangulation where more than one person investigates the same phenomena; theory triangulation where alternative or competing theories are applied to the situation, and finally, methodological triangulation.
Methodological triangulation appears the most complex, comprising ‘within method’, whereby the same method is used on different occasions, and ‘between method’ where different methods are employed in relation to the same subject. To this, he adds a further possibility: checking interpretations with the participants to ensure accurate representation of views and attitudes.
Faulkner’s (1982, in Wellington, 2000)notion of ‘triads’ is extremely similar in that he proposes enquiry should rest on three ‘legs’, each of which represents a separate method of data collection; for example interviews, observation, and scrutiny of documents.
The concerns of what constitutes truth, validity and reliability in interpretive research and to what extent research findings can be applied to alternative situations (generalisability) continue to be debated between epistemologies and even within interpretive research. While one may argue the advantages and disadvantages of this multimethod approach, as one can with all methods of forming validity and reliability, it really becomes a debate on the epistemological stance of those participating in the debate. Silverman (1985, in Cohen et al. 2000) suggests that triangulation is positivistic in stance and “that this is most clearly exposed in data triangulation as it assumes that a multiple data source (concurrent validity) is superior to a single data source or instrument”.
5. Confusion for Novice Researchers?
“Simplicity and rigour” are preferable to “an illusory search for ‘the full picture’”. Silverman (2000: 100)
Good case study research requires at least 6 different types of information. Yin (2003)
For the novice researcher, deciding on methods of enquiry can be confusing. According to Mason (1995), in qualitative case studies, methodological triangulation aids data corroboration and Yin (2003) confirms that good case-study research requires at least six different types of information. This implies that we may obtain a ‘true’ understanding of the situation by seeking the intersection of, for example, a range of interview and observational data though a process of ‘overlaying’. However, Silverman (2000:100) suggests “simplicity and rigour” are preferable to “an illusory search for the ‘full picture’”.
6. Painting a rich picture “No single method holds the key to truth.”Denscombe (2003)
“The combination of multiple methodological practices…adds rigour, breadth, complexity, richness and depth to an inquiry.” Flick (2002: 229)
Fielding and Fielding (1986) clarify the position somewhat; although multiple sources of information cannot give ‘objective truth’ they can be more effectively applied if the researcher follows a set of guidelines: begin with a theoretical perspective (for example, constructionism); chose methods and data which will provide an account of structure and meaning within the context under examination. For Webster and Mertova (2007) similarly, triangulation does not add to the validity of a narrative approach to research; they believe that the real test of validity is that readers find the account ‘believable’. As Flick (2006) states, triangulation is not a tool of validation, but an alternative to it.
7. Slide 7 provides an example of rich data collection
This allows for different lines of enquiry and provides qualitative and quantitative data.
8. A richer picture - Crystallization?
“Validity for postmodernist texts is not the triangle…the central imagery is the crystal…which combines symmetry and substance with an infinite variety of shapes, substances, transmutations, multi-dimensionalities, and angles of approach.” Richardson and Saint-Pierre, 2005, in Denzin and Lincoln, 2005
Richardson and Saint-Pierre talk of “CAP (creative analytical processes) ethnographies” (p2005:962). In this postmodern view, knowledge is seen as fragmentary, situated, and influenced by the self; there are far more than three sides to any story. Rather than working with a fixed, two-dimensional model, hoping to find a central ‘truth’, they use the metaphor of a crystal which acts as a prism. Here there are multiple viewpoints; what we see depends on our angle and there can be no single truth. This multi-faceted structure produces deep, yet complex, and, paradoxically partial understanding. Taylor (1996:44) explains the paradox is that “we know more and doubt that we know”, but has no doubt that the importance of “struggle, ambiguity and contradiction” within the data that “provides us with a deepened complex, thoroughly partial understanding “ of that which we research.
A crystalline approach, according to Richardson (1997, in Denzin and Lincoln, 2003) allows for shifts, changes and alterations in focus without being “amorphous” (p.280). This notion of research as an interweaving of “discovery, seeing, telling, storying and representation” (p.280) perhaps represents more clearly the process an interpretive researcher experiences as she struggles to present a representative rather than ‘accurate’ picture of the subject.
· Provides a range of views and perspectives
· Provides a wider picture
· Allows findings to be corroborated or questioned
· May increase validity
· Is a pragmatic strategy (Schatzman & Strauss 1974)
· Enables questionnaire data to be verified through interview
· Allows qualitative and quantitative data to be combined
According to Flick (2006) knowledge obtained through qualitative methods may be further grounded by “systematically extending and completing the possibilities of knowledge production” (390). Where a variety of methods are used to build a rich picture, discrepancies and inconsistencies may occur, but, as Bryman (1998) and Matheson (1988) suggest, these should be expected. The researcher should not question if what they have found is ‘right’ but undertake further probing to explain the discrepancies.
"In practice, triangulation as a strategy provides a rich and complex picture of some social phenomenon being studied, but rarely does it provide a clear path to a singular view of what is the case. Because of the predominance of the assumption that triangulation will result in a single valid proposition, we look for convergence of evidence and miss what I see as the greater value of triangulating. More accurately, there are three outcomes that might result from a triangulation strategy….convergence, inconsistency, and contradiction". Mathison (1988: 13, 17)
· Each data set may be under analysed
· Overlaying data is a complex process
· Fragments of data may be used out of context
· May lead to over generalisation
· ‘Verification’ may not be accurate – people do not necessarily tell the ‘truth’
Unlike navigational triangulation, in which a ‘true’ point can be established, this is not possible in interpretative research. Different perspectives might point in similar directions, but are unlikely to converge at a point of reality. Hammersley and Atkinson (1983) caution that collection of multiple sources of data can create, rather than solve, problems in reaching understanding; a common problem, for example, when working with multiple sources is to under-investigate the original dataset.
11. Opportunities for Reflection
· Can triangulation ever provide an accurate picture of what has happened?
· Can triangulation add validity to data?
· By examining multiple data sets is there a danger of overlooking an important element?
· Is the crystal a more effective metaphor?
We would be delighted if you responded to these discussion points on The Hive. Thanks, Annie & Sharon
Bryman, A. (1998) Quality and Quantity in Social Research. London: Unwin Hyman.
Cohen, L., Manion, L. and Morrison, K. (2000) Research Methods in Education (5th Ed.) London: RoutledgeFalmer
Cresswell, J. (2007) Qualitative Inquiry & Research Design. Thousand Oaks: SAGE.
Denscombe, M. (2003) The Good Research Guide (2nd Ed.) Buckingham: OUP.
Denzin, N. and Lincoln, Y. (eds.) (2005) The SAGE Handbook of Qualitative Research. Thousand Oaks. SAGE.
Denzin, N. and Lincoln, Y. (eds.) (2003) The Landscape of Qualitative Research (2nd Ed.) Thousand Oaks. SAGE.
Eisner, E.W. (1998) The Enlightened Eye: Qualitative Inquiry and the Enhancement of Educational Practice. New Jersey, USA: Prentice-Hall.
Fielding, N. and Fielding, J. (1986) Linking Data. London: SAGE.
Flick, U. (2006) An Introduction to Qualitative Research. London: SAGE.
Mason, J. (1996) Qualitative researching. London: SAGE.
Mathison, S. (1988) Why Triangulate? In Educational Researcher, 2.
Miles, M. and Huberman, A. (1984) Qualitative Data Analysis. London: SAGE.
Silverman, D. (2000) Doing Qualitative Research. London: SAGE.
Silverman, D. (2006) Interpreting Qualitative Data. (3rdEd.) London: SAGE.
Taylor, P. (1996) Papers presented at the Institute of Drama educational Research – July 1995 at Griffin University Nathan Qld Australia. Australia: Routledge
Webster, L. and Mertova, P. (2007) Using Narrative Enquiry. London: Routledge.
Wellington, J. (2000) Educational Research: Contemporary Issues and Practical Approaches. London: Continuum.