In qualitative research, data triangulation means approaching a question from multiple perspectives.
It involves using more than one data source or method to investigate a theory or corroborate a finding.
For something with such a mathematical-sounding name, it’s a pretty simple concept. But it’s a powerful one, too. Ask a group of five boaters why their watercraft sank in the sea and you’ll get five slightly different stories. Each individual tale holds the bias of its teller. Taken together, though, the individual accounts form a deeper and more accurate picture of what went wrong.
This logic applies to qualitative research. Since qualitative data isn’t as cut-and-dry as quantitative data, you need more than one perspective, data type, and method to shore it up.
With triangulation, you give the results a stronger platform to stand on.
When to Use Data Triangulation with Qualitative Data
If you ask me, there’s never a bad time to do data triangulation when you’re working with qualitative data.
From a researcher’s point of view, triangulation can help you verify key details and strengthen your findings—and your argument.
And if you’re someone who uses qualitative data to inform your business decisions, gathering data from multiple sources is smart. It can help you make critical choices with a level of confidence you wouldn’t (and shouldn’t) have if you relied on just one source.
In a 2014 journal article published in the Oncology Nursing Forum titled, “The use of triangulation in qualitative research,” authors Nancy Carter, et al., lay out two views of data triangulation.
They argue that it’s both “the use of multiple methods or data sources in qualitative research to develop a comprehensive understanding of phenomena” and “a qualitative research strategy to test validity through the convergence of information from different sources.”
Put into simpler words, qualitative data triangulation helps us:
- Gain a fuller understanding of a research topic or argument
- Put our research to the test by comparing and contrasting it with different sources on the same topic
Well-rounded arguments benefit everyone, all the time.
So if you’ve been on the fence about whether to draw sources from more than one place or use more than one method in your research, consider this your sign.
Do it.
How to Do Qualitative Data Triangulation
Before you begin running qualitative data triangulation, it’s helpful to know the four types, as outlined by Carter et al.:
- Method triangulation: Using two or more methods to gather data
- Investigator triangulation: Relying on the involvement of more than one researcher to collect or analyze the data
- Theory triangulation: Analyzing the data using different theoretical perspectives
- Data source triangulation: Drawing information from multiple data sources, including those from different times, places, and people
Knowing which type of triangulation to focus on can be tricky.
We’ll explore each type in more detail and suggest questions to ask yourself when you’re tackling the beginning of the triangulation process.
Method triangulation
Method triangulation means using a variety of research methods to study the same topic. In qualitative research, the most common data-gathering methods are:
- Interviews
- Focus groups
- Observation
- Open-ended surveys or questionnaires
So, in qualitative research, method triangulation means gathering data using at least two of these methods.
But method triangulation can also mean looping in non-qualitative forms of data collection, like demographic information or responses to closed-question (yes-or-no and/or multiple-choice) surveys. Since we’re focusing on qualitative data here, though, we’ll save mixed-method triangulation for another day.
Right now, our lens is firmly focused on qualitative research.
Use method triangulation when:
- Leaning heavily on one method will only give you a partial view of your research question or topic. Say you’re sending out a survey to find what people find most stressful about in-person grocery shopping. You get some interesting answers, but you want to dig deeper. You decide to send a group of shoppers into a store with cash and a grocery list to observe their lived, in-the-moment experiences. Along with the survey responses, you now have a more complete picture of common grocery-shopping stressors.
- You need to cross-validate findings from different methods. Imagine you’re doing in-depth interviews with employees to understand their job satisfaction. At the same time, you want to see if these self-reported experiences align with behaviors you can observe during the workday. You decide to pair the interviews with direct workplace observations. The goal? To see whether the employees’ daily actions match their satisfaction levels. If you see any inconsistencies, you can go over them with the interviewee for a more accurate picture of their experience.
Investigator Triangulation
The goal of investigator triangulation is to have more than one researcher (or team of researchers) analyze the same set of data. Like a peer review for a scholarly journal article, investigator triangulation helps reduce bias. This, in turn, strengthens the credibility of your research.
But you have to be careful not to invite researchers with your same opinions and biases to participate in this type of triangulation. You don’t want them to confirm everything you’ve researched. You want them to read it line by line, grappling with the information and pushing you to see it in a new light.
Reach out to people in different—but related—fields. Invite them to collaborate by analyzing your research and engaging with it from their own viewpoints. Listen carefully to what they have to say—don’t just dismiss it because you don’t see things the same way.
This is how you’ll get the most well-rounded analysis of your qualitative research.
Use investigator triangulation when:
- The research involves subjective interpretations or complex topics. If your data could be viewed differently depending on the researcher’s perspective, then you need fresh eyes to look at it. If you’re studying how patients perceive care in a hospital setting, for example, one researcher might focus on emotional aspects like empathy. Another might not consider this at all and instead focus on how efficient the care is. Both perspectives are important to your research.
- The topic is sensitive or controversial. If you feel like your data is going to ignite a firestorm of controversy, you need extra eyes. Lots of them. The more scrutiny before the data gets published or used to inform a decision, the better. You want minimal personal bias—and maximum credibility.
Theory Triangulation
With theory triangulation, you aren’t using different data collection methods or bringing in researchers with unique viewpoints.
Instead, you’re changing the lens through which you see the data.
This approach challenges researchers to set aside their original theories for analyzing information. It invites them to use at least one additional, theoretical perspective when they sit down to interpret the data.
Researchers usually use theory triangulation when their topic spans more than one discipline. If you were studying human grocery shopping behavior, for instance, you could analyze the results through three lenses:
- Psychological: Study how individual decision-making processes, emotions, and cognitive biases affect shopping choices. Does impulse buying play a role? What about decision fatigue?
- Sociological: Examine the influence of social factors. Do cultural norms, peer pressure, or family dynamics affect shopping habits? Does social class impact purchasing behavior? If so, how?
- Economic: Analyze the shoppers’ behavior through the lens of cost-benefit analysis, budget limits, and price sensitivity. How do incentives like discounts or promotions influence purchases—or not?
Basically, theory triangulation pushes you to consider things from viewpoints you hadn’t before. And it can make the results a lot meatier than if you relied on a single theory.
Use theory triangulation when:
- You are studying a complex topic that could use a few different theoretical perspectives to be understood. Say you’re looking at the factors behind employee motivation. You might use psychological theories like Maslow’s Hierarchy of Needs to explore intrinsic motivators. Next, you could use economic theories to study how external factors like financial incentives influence performance.
- You want to compare or test the validity of multiple theories to see which framework best fits your data. Let’s imagine you’re studying educational outcomes. You could compare Constructivist Theory (focused on how students build knowledge) with Behaviorist Theory (focused on reinforcement and discipline), to see which one better explains student success.
Data Source Triangulation
With data source triangulation, your goal is to gather data from at least two sources, but probably more than that.
What does this look like in qualitative research?
It might mean gathering data from:
- Archival records
- Textual analysis of policy/legal documents
- Social media content
- News articles, blogs, and other media content
- The comments section of any forum, website, or blog
- Case studies
- Literature and artwork
The point of data source triangulation is to study one topic using these diverse data sources. (If you want to pull from quantitative data sources like web analytics and public databases, you can do that too.)
This is essentially another way to study your research question from multiple perspectives. But instead of a group of different researchers or a set of theories from multiple disciplines, those differing data sources are the other perspectives.
Use data source triangulation when:
- You need to collect data from different sources to gather perspectives on a topic—without direct interaction. For example, say you’re studying public perceptions of climate change. You want to capture the raw, unfiltered feelings and opinions behind this fraught topic. What better place to go than social media discussions, news articles (and their comments sections), and government reports? You’ll get tons of rich, probably brutally honest data without ever making your presence as a researcher known.
You want to validate findings across existing qualitative data sources. Let’s say you’re studying stigmas on mental health issues. You’ve already used methodological triangulation to gather qualitative data from interviews and surveys. Now, you want to compare this data with themes from online forums, blog posts, and personal memoirs. The data found in these sources can help validate your findings—or bring up new questions and interesting discrepancies to explore.
