Triangulation

A topic in research methodology

This is now applied to approaching the same research question using several sources of evidence.

"The term triangulation derives from the method of locating a beacon by finding its bearing from different locations: the several directions give limited information individually, but collectively establish the source of the signal. By analogy, the term is also used to describe a common approach to data collection in the social sciences (such as education). Triangulation is considered to be one of the criteria that marks out high quality studies …"

Taber, 2008: 70

Triangulation requires redundancy in the information available.

Read about data redundancy for triangulation

There are different kinds of triangulation.

Methodological triangulation

combining different research techniques to address the same question:

• e.g. combining observation data with interview data
• e.g. combining life-history interviews with documentary analysis

"In methodological triangulation, different sources of data are collected…

For example, a teacher may be interviewed about her teaching, but the interview data is compared to that from other interviewees (e.g. her students and colleagues), observations of her teaching, examination of the comments she makes on student work etc…

The assumption is that an informant, even an honest informant, will have biases, and will have access to only some of the relevant perspectives, and so sound conclusions can only be drawn from evidence that is corroborated from several data sources. Triangulation is primarily used, then, due to considerations of validity – for the purpose of raising the 'trustworthiness' of interpretations"

Taber, 2008: 70-71

Data triangulation:

combining different data sources, obtained by the same research techniques

• e.g. interviewing different informants;
• e.g. interviewing the same person in different contexts, or at different times

Alternatively…

Sometimes triangulation may be understood in terms of using multiple researchers; or analysing data according from multiple theoretical perspectives (analytical pluralism).

Read about analytical pluralism

(Multiple observers/analysts are often used in studies that use ratings scale or classification systems/typologies, to check that different observers are applying criteria in the same way. Usually in these cases only a sample of observations/data are analysed by the researchers to confirm a high level of agreement between the researchers. We might consider the double-rating of the sample as a form of triangulation.)

Read about inter-rater reliability

Source cited:

Taber, K. S. (2008) Of Models, Mermaids and Methods: The Role of Analytical Pluralism in Understanding Student Learning in Science, in Ingrid V. Eriksson (Ed.) Science Education in the 21st Century, pp.69-106. Hauppauge, NY: Nova Science Publishers.