Doing Triangulation and Mixed Methods

Uwe Flick, Author

• Publisher: SAGE Publications Ltd
• Year: 2018
• Print ISBN: 9781473912113
• eBook ISBN: 9781526426222
• Edition: 1
• Topic: Social Sciences
• Subtopic: Social Science Research & Methodology
• Index: Social Sciences

This book shows you not just how to use triangulation as a strategy of quality management, but also how to use it as an approach to designing and doing qualitative research in a more comprehensive way. Flick links triangulation with current debates about using mixed methods, and outlines their potential for extending qualitative research, addressing questions such as how such research can benefit from integrating quantitative (mixed methods), or from working more generally with more than one approach (triangulation).

Chapter One Why triangulation and mixed methods in qualitative research?

Contents

• Triangulation and mixed methods in qualitative research 2
• Triangulation in the history of qualitative research 2
• Triangulation in the discussion about mixed methods 5
• What is triangulation and what is not? 6
• What is mixed methods and what is not? 7

Chapter objectives

After reading this chapter, you should know:

• the background of triangulation;
• why it was originally introduced in social research;
• what is triangulation and what is not;
• the background of mixed methods;
• why it was originally introduced in social research;
• what is seen as mixed methods and what is not; and
• what is necessary for a first orientation about the issues unfolded in the rest of the book.

Triangulation and mixed methods in qualitative research

Triangulation and mixed methods are concepts that refer to a similar idea but make it concrete in differing ways. Both are mainly recognized for combining more than one methodological approach in a study. Mixed methods are focused on combining qualitative and quantitative methods. Triangulation is broader in the kinds of methods that are combined. It is often used as a combination of several qualitative methods. At the same time, it has a broader focus for what is combined – for example, several theoretical approaches or several kinds of data. Triangulation and mixed methods are mostly discussed in alternative and separate discourses, although both discourses can profit from each other. Both concepts can be helpful for extending qualitative research designs (see also Flick, 2018b) for several purposes ranging from extra knowledge to confirming research findings.

Triangulation in the history of qualitative research

Triangulation is a concept that is often taken up in qualitative research when issues of quality are discussed (see also Flick, 2018a). The major link between triangulation and quality of qualitative research is that triangulation means extending the activities of the researcher in the process beyond what is ‘normally’ done – for example, by using more than one method. However, triangulation is not limited to promoting or assuring the quality of qualitative research, but it is also used to extend the range of insights and knowledge produced in a qualitative study. Both of these aims also apply to mixed methods research, which focuses on combining qualitative and quantitative research and thus extending the research approach and making it more valuable.

The different ways of extending research activities with the aim of extending knowledge will be unfolded in this and the following chapters in some detail. To do this, we will address the theoretical and conceptual basis of triangulation in this chapter before the use of different forms of triangulation is discussed in the following chapters. Different aims, and sometimes myths and reservations are linked to triangulation. Sometimes it is discussed when qualitative research is combined with quantitative approaches in order to give its results more grounding. In general, discussion of triangulation in qualitative research began in the 1970s when Norman Denzin (1970) presented a more systematic conceptualization of triangulation. If we go back a little further in the history of qualitative research, we will find that many of the studies seen as classic studies in qualitative research have not used the concept explicitly but were run according to the principles and practices of what is now discussed as triangulation. We may also find that such practices of triangulation can be seen as a feature of qualitative (also including the use of quantitative) research, as some examples may show.

If we go back to early examples of studying social problems in the history of social research, we find works such as the ‘Polish Peasant’ or the ‘Marienthal’ study (see below). Here, we see the combination of several methodological approaches as a starting point for unfolding problems such as unemployment or migration. Combining perspectives on a social problem, William F. Whyte’s (1955) classic ethnographic study of a street gang in a major city in the eastern USA in the 1940s offers, on the basis of individual observations, personal notes and other sources, a comprehensive picture of a dynamic local culture. Through the mediation of a key figure Whyte had gained access to a group of young second-generation Italian migrants. As a result of a two-year period of participant observation he was able to obtain information about the motives, values and life-awareness and also about the social organization, friendship-relations and loyalties of this local culture.

The study by Marie Jahoda, Paul Lazarsfeld and Hans Zeisel (1933/1971), Marienthal: The Sociology of an Unemployed Community, is one of the classic studies in qualitative research (see also Fleck, 2004). Here, psychological coping with unemployment in a village was studied in the late 1920s after the main employer of its inhabitants went bankrupt. The result is the elaboration of the leitmotif of a ‘tired society’ as a condensed characterization of the attitude towards life and the day-to-day practices in the village and of different types of practices in reaction to the unemployment (for example, the ‘unbroken’, the ‘resigned’, the ‘desperate’ and the ‘apathetic’). Jahoda (1995, p. 121) has summarized the methodological procedures leading to these insights in combining approaches, such as qualitative and quantitative methods, objective facts and subjective attitudes, observations of current practices and historical material and observations with planned interviews. These principles include linking different methodological approaches (qualitative, quantitative, interviews and observation). At the same time, we find different methodological perspectives (objective facts, subjective attitudes, current and historical issues). In describing the study (1933/1971), the authors list the data they used: cadastre sheets for about 500 families, life histories, sheets for documenting the use of time, protocols, school essays, different statistical data, and historical information about the village and its institutions. Accordingly, Lazarsfeld (1960, p. 14) has made the link between qualitative and quantitative data and strategies a principle, at least for this study. According to Lazarsfeld (1960, p. 15), ‘three pairs of data’ were used for the analysis: ‘natural sources’ (statistics of library use) and data that were collected for research purposes (sheets of time use); ‘objective indicators’ (e.g. health statistics) and subjective statements (interviews); and ‘statistics and empathic descriptions of single cases’.

Another early example of triangulation of verbal and visual data is the study by Gregory Bateson and Margaret Mead (1942). A remarkable feature of the study is the empirical approach of producing and analyzing more than 25,000 photographs, masses of filmed material, paintings and sculptures on the one hand and using ethnographic conversations about this material on the other hand. Also, in The Polish Peasant in Europe and America by Thomas and Znaniecki (1918–20), different sorts of data were combined: ‘undesigned records’ as well as an exemplary life history produced by a participant for the study. Finally, Morse (2003, p. 190) sees the work of Goffman (e.g. 1989) as an example of applying triangulation without using the term. These examples show that the use of different sorts of data was characteristic of many classic studies at the beginning of qualitative research.

Barney Glaser and Anselm Strauss and their approach of discovering grounded theories were at the core of the renaissance of qualitative research in the 1960s in the USA and in the 1970s in Europe. Not only the methodological works, but also the studies (e.g. Strauss et al., 1964), are influential and instructive. Again we find different hints for the use of what was later named triangulation. Glaser and Strauss suggest the use of different types of data, what they call ‘slices of data’ collected in ‘a multi-faceted investigation’ (1967, p. 65). They also suggest using many different sorts of data, whereas Strauss et al. (1964, p. 36) advocate employing different observers in order to increase the reliability of observations that were made independently of each other and comparing them.

These examples may show that the triangulation of data sources, of methods and of researchers has a long tradition in various areas of qualitative research, even if the term was not (yet or always) used. These examples also demonstrate that, in the tradition of these studies, triangulation as an empirical approach to fields and issues was employed as an instrument assessing empirical results as well as a way to gain more insights and knowledge in the research.

Triangulation in the discussion about mixed methods

When the recent discussions about mixed methods started, the existing approach of triangulation was reduced to a specific understanding. Here, the confirmation of results from one approach by those of another approach is regularly seen as its major task (see e.g. Bryman, 2004). Greene et al. (1989) have introduced this limited view in the discussion by referring to two major sources. On the one hand the concept of ‘multiplism’ as outlined by Cook (1985) is suggested for when it is unclear which is the best of several methodologies for a study, to triangulate the most useful or promising ones. On the other hand, referring to Mark and Shotland’s (1987) suggestions of purposes for multi-method designs: (a) triangulation, which seeks convergence of findings; (b) bracketing, which seeks a range of estimates on the correct answer (or triangulation with a confidence interval); and (c) complementarity (Greene et al., 1989, p. 257). With the background of these sources, the field that covers triangulation and mixed methods as ways of engaging multiple (methods) approaches is charted as follows: triangulation is limited to obtaining ‘convergence, corroboration, correspondence of results’ and juxtaposed against complementarity (aiming at elaboration or enhancement), development (aiming at further sampling) and expansion (aiming at more breadth and range) by using multiple methods (see Greene et al., 1989, p. 259 for details).

In the following chapters of this book, we will discuss: first whether charting the field of multiple approaches in this way will underestimate the potential of triangulation in several ways; second, how far it leads to some of the current shortcomings in the discussion about mixed methods (see Flick, 2018c); third, whether it gives up the option of using triangulation as a concept to overcome some of the challenges mixed methods research is currently facing; and fourth, whether it might reduce the potential of using multiple methods within the context of qualitative research. For this purpose, I will start with a short outline of both concepts – triangulation and mixed methods – in this chapter, which then will be unfolded in more detail in the following chapters.

What is triangulation and what is not?

Why triangulate?

Triangulation attracted interest as a methodological strategy for a long time in social research, in evaluation, and qualitative research in particular, as an obligation to use combinations of methods, of data, and of researchers for improving the validity of a study and to ‘withstand critique by colleagues’ (Mathison, 1988, p. 13). This attraction was linked for some time to the combination of convergence of results, their validation and legitimation of research by applying triangulation. However, triangulation very soon became the object of a more or less fundamental critique. Put simply, the concept of triangulation means that an issue of research is considered – or in a constructivist formulation is constituted – from (at least) two points. Normally, the consideration from two or more points is materialized by using different methodological approaches (see Chapter 2). The concept of triangulation was imported from land surveying and geodesy, where it is used as an economic method of localizing and fixing positions on the surface of the earth (see Blaikie, 1991, p. 118). The definition used in this context sees triangulation as a method of locating a point from two others a known distance apart (Clark, 1951, p. 145).

In a more metaphorical sense, Campbell and Fiske (1959) and Webb et al. (1966) introduced triangulation into general methodological discussion in the social sciences. At that point, the idea was already that the issue under study is also constituted by the methods used to study it. At that time, a rather negative reading was dominant: that the issue is possibly biased by the methods that are used and that results have to be seen as artifacts. The leading question was whether ‘a hypothesis can survive the confrontation with a series of complementary methods of testing’ (Campbell and Fiske, 1959, p. 82). This led to considerations of how to prevent such a bias, and ‘unobtrusive’ and ‘nonreactive measurement’ (Webb et al., 1966) were stipulated. One strategy was the combination of different measurements and methods – in a ‘multitrait–multimethod matrix’ (Campbell and Fiske, 1959). In this context, the metaphor of triangulation is imported ‘from navigation and military strategy that use multiple reference points to locate an object’s exact position’ (Smith, 1975, p. 273) to the social sciences.

For a better understanding of the concept of triangulation, it might be helpful to see what is not meant by it. In ...

Table of contents

• Cover
• Half Title
• Publisher Note
• Title Page
• Copyright Page
• Contents
• Sidebar List
• Illustration List
• Table List
• Editorial introduction
• About this book
• Chapter One Why triangulation and mixed methods in qualitative research?
• Chapter Two What is triangulation?
• Chapter Three Methodological triangulation in qualitative research
• Chapter Four Triangulation in ethnography
• Chapter Five What is mixed methods research?
• Chapter Six Designs, methods and data in mixed methods research
• Chapter Seven Triangulation as a framework for using mixed methods
• Chapter Eight How to use triangulation and mixed methods in qualitative researchPractical issues
• Chapter Nine Sensitive use of multiple methodsQuality, writing, and ethics
• Glossary
• References
• Index