Triangulation (social science)
Triangulation is a research strategy in empirical social research, various methods or perspectives are applied to the same phenomenon when or different data for the study of a phenomenon can be used to compensate with the strengths of each one procedure, the weaknesses of each other. The aim is usually to achieve higher validity of the research results and to reduce systematic errors. However, a number of authors believes triangulation would only provide a richer, but not necessarily allow valideres picture of empirical reality. A minority of researchers, mainly from hermeneutic research traditions, rejects triangulation of epistemological reasons completely.
Although triangulation has an affinity for mixed -methods designs, it is now mostly in the qualitative social research, within the paradigms of the past 40 years, the greatest development of triangulation has occurred. History of ideas, but it is rooted in the quantitative social research.
- 4.1 Choice of metaphor
- 4.2 problems with conflicting results
- 4.3 Ontological and Epistemological Questions 4.3.1 Realism accusation
- 4.3.2 Alternative to or confirmation of Validitätsgedankens?
History
The combination of different methods and data in social research has been done in the 19th century, for example, in Lenin's 1898 erschienenem The Development of Capitalism in Russia. Such Combine was throughout the first half of the 20th century - for example, in the classic study The unemployed of Marienthal - practiced for the common use today: metaphorical term " triangulation " was imported but only from the 1950s from geodesy in the social sciences. At the beginning of the 1960s were especially more quantitative -oriented researchers such as Paul Lazarsfeld advocate of combinations of methods, although qualitatively working sociologists such as Howard S. Becker also " documents of different types " due to their higher credibility preferred, without mention but this " triangulation ". It was not until the late 1960s, won triangulation by the rise of Grounded Theory But as explicitly pursued research strategy and the general methodological literature increasingly important. A typological systematization of various forms of triangulation emerged only in the 1970s. From the middle of that decade the triangulation found its way into many standard textbooks on social research, however, that systematic methodological approaches have been taught without. From the mid- 1980s created the first systematic reviews of triangulation, which has resulted in parts of the research community to shift their legitimacy grounds off the usual Validitätsbegriffe. As part of since the beginning of the new millennium increasingly funded integrated methods training in the social sciences, triangulation gained further importance. However, strong ideological contradictions between quantitative and qualitative research have prevented a more rapid increase in the use of triangulation.
Dissemination
Parallel to the importance of qualitative methods in the various social science disciplines, the triangulation is especially prevalent in their strongholds such as ethnography; in more quantitatively oriented disciplines such as economics, however, it is more rare. Discipline Cross- triangulation is rather a desideratum in textbooks as a common practice in the research program, although, for example, in 2004 in a poll of British educational researchers disclosures 71% of all respondents said they would rely on triangulation. By contrast, however, there is an empirical finding that triangulation in the field of business administration in the 1990s, declined in comparison to the 1980s. With regard to the fields of research triangulation is especially often found in nursing, health, tourism and educational research.
Methodology
Triangulation is today, in addition to sequencing, are applied to the qualitative and quantitative research strategies in sequence, and hybridization, in which the mixing of methods can be identified analytically little more, a core area of combination methods in the social sciences. However, it should be noted that due to the popularity of the term triangulation is by no means a clearly circumscribed research program more.
Legitimacy
In general, to give research a higher plausibility and credibility triangulation. The basic idea is that a series of different each fault and problem-ridden measurements or analyzes in the aggregation a more valid analysis would produce, by the weaknesses of one approach would be offset respectively by the strengths of others, because such errors, problems or measurement inaccuracies were revealed. However, matching triangulation results are not necessarily associated with a high ( external ) validity, instead Triangulationskonsensus is necessary but not sufficient condition for high validity. More recently, also back from a number of researchers from traditional aspects of validity and now only suspect that triangulation not necessarily more valid, but would still achieve richer results. A frequently mentioned special form of this reasoning is the reference to the problem of the micro and macro levels theoretically and empirically to connect with each other, which could only be done by the application of various methods and theories in the eyes of some authors. Although the latter researchers more represented in the tradition of relativism and former rather a critical realism, both traditions see the frequently quoted aphorism Triangulationsliteratur by Richard Levins again:
" [O ] ur truth is the intersection of independent lies. "
This epistemological legitimation of triangulation are facing more pragmatic reasons, in which researchers rely on new methods because they have achieved with the previously used methods forecast results have been found to be too bad.
Typology
Norman K. Denzin 1970 developed the most widely used today and legitimacy typology of triangulation. He thereby distinguishes four types of triangulation: data triangulation, Forschertriangulation, Theorietriangulation and methodological triangulation.
Data triangulation
When data triangulation data is used from different sources or different types of data from the same source, so as to equalize the various different data Biase material. Particular, it is intended to collect data from people in different social positions. In a study on language courses for example, students, teachers and the administrative staff can be questioned, since each of these sources probably have a different approach and perspective to the language courses. Alternatively, or in conjunction with this approach, for example, experimental data, survey data, or from a participant observation of the various groups of people could be gathered. The various data in turn can be present in different shapes and can be evaluated, for example interview data as transcript. (In different transcription forms ) or as a video or audio files
Special forms of this kind of triangulation are the time and Ortstriangulation, are collected in the data at different times or in different places, so as to eliminate idiosyncrasies of the historical context.
Forschertriangulation
Another possibility is that several researchers to analyze the data. The participation of more than one researcher to the ( cognitive ) conflict of interest that exists when one and the same research findings formulated theoretically and empirically, defuse. Furthermore, it is assumed that different people will follow different theoretical approaches and thus at the same time made a Theorientriangulation (often implicit). This is sometimes even so far as to deliberately researchers are recruited, their social background, such as their gender or their social class varies, because that is considered to be important for their access to data. Finally, it is hoped that in qualitative social research from the Forschertriangulation also that various researchers bring different skills in data analysis. Particularly common Forschertriangulation is applied in qualitative social research at participating methods such as participant observation, as well as traditional quantitative methods such as content analysis using this strategy and validate codebooks using the intercoder reliability.
Denzin recommends it, not as usual to let students or graduate students perform the encoding work, but to resort to experienced qualitative researchers.
Methodological triangulation
The triangulation of methods is the most widely used method of triangulation, though the use of different methods, of course, often the use of different data, and therefore data triangulation implies. In the late 1950s and early 1960s, when triangulation rather a domain was quantitatively post positivist research, especially in psychology, this form of triangulation was limited to the use of different measurement models or operationalization, initially only checked the reliability of different operationalizations and should be improved and then later more valid measurements were sought. Today, the field of methodological triangulation is taken much further, using not only different measurement methods, but various empirical research methods are used.
Often a distinction is made in the triangulation of methods with Norman Denzin between a triangulation within a method and between methods. Triangulation within a method, the same method is applied to a variety of ways. In quantitative surveys this can be done for example by the use of different indicators for measuring the same construct; in qualitative research this can be done, for example, in ethnographic studies using the reference to several different study groups.
The combination of different methods come in two basic ways: On the one hand data from various sources, on the other hand, the same data in a combination of methods with different methods can be evaluated. Alternatively, the same data can be converted, usually by qualitative data is converted by encoding into quantitative data.
If one chooses a combination of methods, so you can combine in Denzin's typology, both within and between methods. The combination within methods means that you can apply the same method in different types, such as occurs for example in the multi-dimensional scaling. Today, however, most combined between methods, and in this case this is usually the combination of one or more quantitative methods with one or more qualitative methods, in fact, this variant of the method triangulation has become so dominant that make authors define it as only form of triangulation. This can be combined with each other in both directions qualitative and quantitative methods. If one begins with a quantitative survey, one can identify specific cases, so-called outliers, for a more detailed investigation using qualitative methods or avoid that in the selection of cases for qualitative studies of the typical " elite bias" striking through the application of statistical techniques. In the opposite direction you can check in larger populations developed using qualitative methods theories through the development and application of appropriate quantitative measuring instruments. One particularly common example of the latter version is the use of quantitative surveys in the series of qualitative interviews.
Such combinations of methods can, however - regardless of the ontological and epistemological assessment of the triangulation itself - problems arise when the methods used imply contradictory to each other ontologies. This has the consequence that the combination of different methods always first presupposes the choice of a particular epistemological research paradigm.
Finally, it is to be observed in the choice of method is always that the methods chosen are congruent with the theories used in each case; Of methods should not be an end in itself.
Theorientriangulation
When Theorientriangulation different theoretical perspectives on the same phenomenon or the same data are used. Denzin enumerated here interactionism, Marxism, phenomenology, feminism, semiotics and cultural studies as a possible theoretical perspectives.
This form of triangulation is probably the most difficult to reach; Some authors even think Theorientriangulation is probably unattainable.
In the 1990s, Valerie Janesick suggested a fifth type of triangulation, the interdisciplinary triangulation ago. Similar to the theoretical triangulation are here approaches of different disciplines used to develop the explanation of a phenomenon.
Type combinations
The different types of triangulation can of course be combined with each other. A particularly well-known example of this type within the ethnomethodology is Cicourels unlimited triangulation involved in the both researchers and research subjects and research assistants in the generation of different types of data material and its evaluation. Here, various forms of data by discussions between researchers and researched are created, for example by using different transcription styles for generating data for an analysis of the same conversations are used.
The role of a computer program
When data triangulation of many authors consider programs for computer-assisted qualitative data analysis ( CAQDAS programs) are helpful because they can organize a number of different types of data; sometimes even an impetus for more widespread use of triangulation is hoped the development of such software. These programs can allow the qualitatively working researchers in terms of data and methodological triangulation to pass it encodes data without loss of statistical programs. The investigation would gain as transparency and systematics. In addition, various collaboration features of CAQDAS enable the comparison of the analyzes of the same data material by different researchers.
It is generally expected from the use of software in the triangulation of methods, an increase of analytical rigor ( " rigor " ) and reliability; Studies will be "deeper" through the use of software and have more details. A minority of authors, however, sees in the use of CAQDAS a tendency to rigidity of qualitative social research that may lead to the deterioration of the research potential of these methods.
An application example of the use of the method triangulation is CAQDAS extracting transcription portions which have been encoded in the same quantitative way.
Assessment
Although triangulation first seems intuitively appealing, it is on closer inspection but not entirely unproblematic. Besides the choice of metaphor, especially forschungspragmatische and epistemological problems posed.
Choice of metaphor
A recurring criticism is the choice of metaphor (the term triangulation comes from geodesy yes ), the two-dimensionality would be for an outdated epistemological worldview.
Problems with conflicting results
On the purely pragmatic level, the question arises, how to deal with Triangulationsergebnissen each other neither confirm nor enrich, but contradict. In practice, researchers are more likely to trust the results achieved by qualitative methods, but rather the emotional involvement awareness of researchers is due to qualitative data for a formal logical reasoning legitimacy; one therefore unacceptable approach. A well-reasoned solution to this problem but there is not today. Alan Bryman recommends then succinctly to make inconsistencies to the starting point of new ideas to improve theory.
Ontological and Epistemological Questions
However, the sharpest reviews by far the triangulation originate from epistemological and ontological considerations. Particularly worth mentioning is the realism reproach.
Realism accusation
Like all methods of empirical social research is the triangulation epistemological premises. In this case, it is a ( naive ) realism, which is primarily accused the advocates of triangulation of constructivist page because triangulation necessarily presupposes the existence of an objective reality. Some authors, including Norman WH Blaikie and Yvonna S. Lincoln, therefore even believe that triangulation successes in practice only within (post) positivist approaches. Other authors, including Clive Seale, have this allegation to the effect that triangulation can be used despite its affinity for realism in other epistemological paradigms. In particular, a pragmatism Deweyscher embossing is likewise well suited as " epistemological partner."
Alternative to or confirmation of Validitätsgedankens?
From representatives of postmodernism have been urged against the fact that triangulation ignores the fact that each method throws a different view of a phenomenon; they propose that instead of triangulation crystallization in the combination of methods before, are to be made in precisely not equivalent results.
Status
Despite all the advantages and disadvantages of triangulation, however, remains to be noted that it is but usually not seriously considered, despite numerous supporters from the majority of the particular postmodernism -leaning social science research.