Applying Triangulation techniques to enhance mixed methods research design and analysis

How do you define Mixed Method Technique:   Selection and Construction of Mixed Method research is a process. Sometimes it is also called as Mixed Method Research (MMR) Designs. The use of the word design has two implications when it comes to mixed method research. The first method focusses on the process of design where design is often seen being used as a verb. The other implication of the use of this word is the result of designing. Both these meanings hold their relevance in mixed method designing. In order to be able to create a strong design as a product, several rules need to be considered carefully. Though obeying the rules does ensure that the design is going to be strong but it certainly contributes towards a robust design. A mixed method design is a contribution of one qualitative and one quantitative component. A more precise definition of mixed method research can be:

It is a type of research in which a team of researchers put together elements of qualitative and quantitative research approaches for the broader objective of breadth and depth of the understanding and analysis to reach meaningful understanding and analysis.

Mixed Method research has similar traits to multimethod research.  The prominent difference being that in multimethod research exclusively multiple qualitative or multiple quantitative techniques are combined where as in mixed method research a combination of qualitative and quantitative techniques are used.

There are several characteristics to mixed method designs that need to be necessarily incorporated in the design process: These are:

• Purpose of mixing
• Theoretical drive
• Timing
• Point of integration
• Typological use
• Degree of complexity

Whenever a researcher uses mixed method research, methodological triangulation will always be used.

What is triangulation in research:

Triangulation in research means the use of multiple data sets, methods, theories, and investigators who would address the research question. It is a very robust technique that can help you to enrich and enhance the credibility and validity of your research findings and dilute the presence of any kind of research biases in your work.

The maximum usage of triangulation is seen in qualitative research but it’s not uncommon in quantitative research. Whenever a researcher uses mixed method research, he must apply the triangulation technique.

 The more precise application of triangulation in different kind research can be understood with the following examples

• Qualitative Research: In-depth interviews are conducted with different kinds of stake holders such as parents, teachers, children etc.
• Quantitative Research: An eye tracking experiment is run through that involves three researchers for the job of data analysis
• Mixed Method Research: A Quantitative survey is conducted followed by selective structured interviews that are qualitative in nature.

Different types of Triangulations in research:

There are four main types of triangulations

• Data Triangulation: data triangulation means the use of data from different times, spaces and people
• Investigator Triangulation: investigator triangulation implies the use of multiple researchers in the collection or analysis of data
• Theory Triangulation: this is used when the researcher is using varying theoretical perspectives in his research
• Methodological Triangulation: this means the use of different methodologies in order to approach the same topic.

Let us understand the four types of triangulations in detail

1. Methodological Triangulation:

As explained earlier, when we use methodological triangulation, different methods are used to approach the same question. It is the most widely used type of triangulation and researchers apply qualitative as well as quantitative methods in one study. Let us try and understand this better with an example

• In your study you use behavioral, survey as well as neural data to get a complete picture of what causes stress amongst people. To do the same, you choose participants to perform team games in behavioral controlled lab experiments and record their observations. You further take up a survey to collect data about the factors that cause stress in their daily lives. In addition to that you perform medical scans on volunteers to know the changes that take place in the neural system to assess the changes taking place in the brain when stress triggers.  This technique works to overcome the problems such as bias that arise with the use of a single technique.

2)Data Triangulation:

In this technique, multiple data sources are used to answer the research questions. The data can be different from each other in context of time, space or the respondents. This depends on the type of study. The example of the same can be seen here

Example: To understand the impact of stress management training and practice, you compile and analyze data from a set of 100 employees in an organization over a period of 6 months. You can even further repeat the experiment with samples in different regions worldwide, receiving similar kind of training. When the data is collected across samples, place and time, the results are expected to be more generalized to other situations.

• Investigator Triangulation:

This technique deploys multiple researchers to observe, collect or analyse the data exclusively.

Example:

 You involve many observers to code the same set of participants to analyse their stress levels of employees in an organisation. What is imperative here is to provide them with training sessions and a pre designed manual to ensure that they code the stress related behaviour of the participants in the same way. As the head researchers you will have to provide recordings to the different researchers to observe the employee’s reaction in different situations and when and how their stress level is depicted in their actions. You will also check and ensure that their code sheets are lined up with each other so that high interrater reliability is there. This technique also helps to control interrater observer bias and also other kind of biases associated with experiments.

4)Theory Triangulation:

Theory triangulation means applying several different theoretical frameworks to approach the research question rather that looking at it from only a single theory perspective.

• An example of theory triangulation can be that there are two competing theory that talk about the factors that increase the level of stress amongst employees. One could be attributed to intrinsic factors affect the stress level of employees and the other could be that the stress level of the employees could be affected by extrinsic factors. Competing hypothesis are created to perform theory triangulation. This kind of technique can help to look at the research question from more than one perspective and also reconcile contradictions if any, in your data.

Application of Triangulation

 The application of triangulation is seen in more holistic perspective for the answer to a specific research question. It also helps to enhance credibility as well as validity. The specific purpose of triangulation can be seen here as

• To cross check evidence: it is necessary to collect high quality data for the purpose of rigorous research. Naturally, when the data is from a single source or from one investigator, he trustworthiness of the data is in question. But the credibility of the data increases when it is from multiple sources. By credibility we mean that how confident one can be that the findings are giving a realistic picture. The more your data coverage agrees with each other, the results will be more credible.
• For a holistic picture: it gives a better and an overall understanding of your research problem. As a researcher, you always risk bias in your research when you rely on a single data source, methodology. Observer bias is seen in the case of a single researcher doing the data collection and analysis. Similarly, using a single theory or one methodology constraints to inherent flaws and limitations to that method. Triangulation is of great help when the researcher wants to capture the complexity of the real-world phenomena. By diversifying into different data sources, theories, and methodologies you can go deeper int looking at the problem for various perspectives and levels.
• To improve validity: It means how accurately a method is measuring what it is supposed to measure. The research validity can be increased through triangulation. Since all the methods have their own strengths and weaknesses, when through triangulation you combine complimentary methods, they account for each other’s limitations. For example, when we compare behavioral observations with survey and neural mechanism, we can identify that behavioural observations sometimes makes participants conscious of their responses because they are being each other. The sorbet method is self-reported so it can lead to observer’ s bias and neural mechanism does not have any of the above limitations but it is not self-sustaining in research. It can be used only as a support technique. By combining all the techniques, one’s flaws can be made into the method’s strengths.

Pros and cons of Triangulation in Research

No research technique is without its own strengths and weaknesses. Same applies to triangulation as well

Some of the advantages associated with triangulation are:

Reduces Bias:

• Triangulation technique surely helps you to overcome the bias that comes with using a single technique or perspective in your research. You will get a holistic perspective to your research question when you apply triangulation technique.
• Builds credibility and validity:

The triangulation technique helps in establishing credibility and validity as different techniques are combined or sometimes different sources of data or theories. With the application of multiple techniques and perspectives, you can know that your findings are more realistic and credible.

• Takes a lot of time:

Triangulation can be extremely time consuming and involves a lot of labor.  Since as researcher you choose to deploy multiple techniques, you will have to juggle different datasets, sources, and methodologies to answer the research question.

• Costly

It involves interdisciplinary team and a higher cost and workload. This can add to the costs extensively and as researcher you will have to strike a balance based on the priority of your research needs and the time frame you have.

• Inconsistency

This technique may sometimes make it difficult to line up the data from different sources, investigators, methods and give a vivid picture. There are times, when, despite all your precautions, your data may contradict with and exhibit inconsistency. This doesn’t make your research worthless or incorrect, however, you would surely have to dig deeper  to understand why there is inconsistency in the data.  But the tray of hope can be that from these inconsistencies and contradictions, new avenues of research can be explored.

Summary and Conclusion

 Triangulation technique in mixed methodologies is used to compare statistical findings in quantitative terms straight from qualitative findings or verify and extend quantitative results with the help of qualitative data. It is a widely used technique to best understand a research problem and acquire additional separate data. It is a pathway to overcome the shortcomings of standalone quantitative data analysis. It corroborates the strengths and non-overlapping limitations such as extensive sample size, generalization, and trends with those of qualitative techniques, such as limited N, details, in depth study.  The procedure of this technique is a one phase system in which the researcher applies qualitative and quantitative methods of equal stature over the same time frame. So basically, the qualitative and quantitative data is analyzed across the same time frame but separately so that the understanding of the research issue can be made better. The researcher attempts to put together both datasets but combining both the outcomes in the interpretation or by the manipulation of the data to allow the convergence during the analysis of two data types.  It is a very useful and profitable tool for beginners in mixed model research. It has an efficient architecture in which both the data types are gathered at the same point during the research phase. Those researchers who have efficiency in both, qualitative and quantitative techniques can be included in this research type as both these types of data needs to be gathered and evaluated separately. Overall, an excellent technique for getting realistic findings and answers to the research question. Despite being time consuming and expensive in execution it is gaining popularity.