Whether it’s Fortune 500 companies or tiny startups, people are recognizing the value of building products with a user-first methodology.
But it’s not enough to merely say “we’re doing research”, it has to be the right UX research. Research that combines richness of different people's experiences and behavioral insights with tangible numbers and metrics. Key to this is an approach called mixed methods research.
Here, we’ll dive into the what and why of mixed methods research and cover a few examples of the approach.
What is mixed methods research?
Mixed methods isn’t some overly complicated practice that’ll take years to master — it simply refers to answering research questions through a combination of qualitative and quantitative data. This might mean running both interviews and surveys as part of a research project or complementing diary study data with analytics looking at the usage of a particular feature.
A basic mixed methods question could be: “What are the key tasks people perform on my website?”. To answer this, you’d look at analytics to understand how people navigate through the page and conduct user interviews to better understand why they use the page in the first place. We’ve got more examples below.
It makes sense: using both qualitative and quantitative methods to answer a single research question will mean you’re able to build a more complete understanding of the topic you’re investigating. Quantitative data will tell you what is happening and help you understand magnitude, while qualitative data can tell you why something is happening. Each type of data has its shortcomings, and by using a mixed methods approach you’re able to generate a clearer overall picture.
When should you use mixed methods?
There’s really no “time to do mixed methods research”. Ideally, for every research question you have, evaluate which qualitative and quantitative methods are most likely to give you the best data. More often than not, you’ll benefit from using both approaches.
We’ve put together a few examples of mixed methods research to help you generate your own UX research questions.
Examples of mixed methods research
Imagine this: You’re on the user research team at BananaBank, a fictional bank. You and your team want to investigate how the bank’s customers currently use their digital banking services so your design team can make some user-focused improvements.
We’ve put together a few research questions based on this goal that would best be served by a mixed methods approach.
Question 1: How does people’s usage of online banking differ between desktop and the app?
The value of quantitative methods: The team can view usage analytics (How many people use the desktop app versus the mobile app) and look at feature usage statistics.
The value of qualitative methods: Interviews with users can answer all manner of questions. For example, the research team might want to find out how customers make their way through certain parts of the interface. Usability testing is an opportunity to watch users as they attempt various tasks (for example, making a transaction).
Question 2: How might you better support people to reach their savings goals?
The value of quantitative methods: The team can review current saving behavior patterns, when people stop saving, the longevity of term deposits and other savings-related actions.
The value of qualitative methods: Like the first question, the team can carry out user interviews, or conduct a diary study to better understand how people set and manage savings goals in real life and what usually gets in the way.
Question 3: What are the problem areas in our online signup form?
The value of quantitative methods: The team can investigate where people get stuck on the current form, how frequently people run into error messages and the form fields that people struggle to fill out or leave blank.
The value of qualitative methods: The team can observe people as they make their way through the signup form.
Mixed methods = Holistic understanding
As we touched on at the beginning of this article, mixed methods research isn’t a technique or methodology, it’s more a practice that you should develop to gain a more holistic understanding of the topic you’re investigating. What’s more, using both types of methods will often mean you’re able to validate the output of one method by using another.
When you plan your next research activity, consider complementing it with additional data to generate a more comprehensive picture of your research problem.
Which comes first: card sorting or tree testing? – We run through why you should use card sorting and tree testing, and which one to use first.
How to write great questions for your research – Whether you opt for a qualitative or quantitative method – or both – learning how to write great research questions is key.