Experienced nurse researchers have generally become polarised in their approach to research through the exclusive use of either a quantitative or qualitative paradigm, one type of data or one theoretical perspective. Within the literature there is general support for the separateness of research approaches, for example qualitative and quantitative paradigms (Duffy 1987). However, in accepting the differences between the approaches, there has been concern among nurse researchers that neither approach in isolation will provide an understanding of human beings and of their health-related needs and nursing care. For more than 20 years there has been an increasing range of nursing literature highlighting the merits of triangulation and demonstrating its use in a wide variety of nursing and midwifery situations (Cowman 1993, Carr 1994, Shih 1998, Carin 2005 and Williamson 2005).

Triangulation is a term originally used in navigation as a strategy for taking multiple reference points to locate an unknown position. Triangulation was first described as research by Campbell 1956 and Campbell 1959 are acknowledged as the first researchers to apply triangulation when they promoted a multiple methods approach. One of the earliest and most recognised authorities on the use of triangulation was Norman Denzin. His first book laid down the earliest understandings and definitions of triangulation; Denzin (1970) discussed triangulation as the combination of multiple methods in a study of the same object or event to depict more accurately the phenomenon being investigated.

However, Denzin (1989) later extended his definition from the measurement of discrete concepts to the level of research design, an approach he terms multiple triangulation. Denzin believed that the purpose of multiple triangulation is to overcome the intrinsic bias of single method, single observer, single theory studies, thus confirming the researcher’s results and conclusions. Similarly Polit and Hungler (1999) presented a more embracing description of triangulation as the use of multiple methods or perspectives for the collection and interpretation of data about a phenomenon to obtain an accurate representation of reality.

In terms of understanding and application Denzin (1970) identified four main types of triangulation (Table 26.1) which were subsequently discussed by Mitchell 1986, Duffy 1987 and Sohier 1988. There have been attempts to distinguish other types of triangulation, such as interdisciplinary triangulation (Janesick 1994) and conceptual triangulation (Foster 1997).

There are several arguments in favour of using triangulation in nursing research, such as the complexity of health services, understanding of which calls for multiple methods and perspectives (Foss & Ellefson 2002). Triangulation, by reconciling the paradigmatic assumptions inherent to qualitative and quantitative methods, can provide rich data whilst limiting the propensity to make oversimplified conclusions. Different types of knowledge are gained through different methods; however, the danger is that a researcher may lose sight of the differences underlying the chosen methods. The epistemology of triangulation should not be portrayed as a mix of two different epistemological positions but, rather, be seen as an epistemological position in its own right. Sim and Sharp (1998) discussed an epistemological paradox involved in seeking to use triangulation as a means of validation in qualitative research. The concept of validation assumes a single objective reality whereas qualitative research is founded on the notion of multiple realities. Denzin and Lincoln (1994) claim that triangulation is not a tool or a strategy of validation, but an alternative to validation.

There are strengths and weaknesses to triangulation (Table 26.2); Bradley (1995) identified that it was initially introduced as a technique to overcome some of the weaknesses of qualitative research in the 1950s and 1960s. I suggest that, by reconciling the paradigmatic assumptions inherent to qualitative and quantitative methods, triangulation has the potential to provide rich and productive data (Cowman 1993).