Triangulating User Intent Using Mixed Methods

Truly understanding your user’s needs lies at the heart of designing successful products and services. When a researcher relies on on only one data source, they can fail to comprehensively understand their user’s motivations, preferences and pain points. Instead, rely on using mixed research methods so that you can paint the fullest picture of your customers. 

Combining Qualitative and Quantitative data

Qualitative research methods like user task testing or interviews are a great way to understand user motivations and beliefs. Reviewing quantitative data like surveys or analytics experiments can back up these insights with statistical significance. When conducting user task testing, I often work with a small sample size (10-20 users) to look for pain points across the site. If users are struggling to find information, I’ll compare my findings with a survey, A/B test or Heatmapping to back up these insights. 

Avoiding Research Bias

One issue with mixed method research is it can lead to research bias where learnings from qualitative data lead the researcher to only look for insights that back this up in quantitative data. To minimize bias, clearly define research objectives at the onset of the project. Secondly, use randomized sampling techniques to ensure a representative user sample. Thirdly, be neutral during data collection and analysis to avoid a subjective interpretation. Lastly, consider diverse perspectives and actively seek contrary evidence to challenge your assumption. 

Mixed Method Research Processes

1. Convergent Parallel - In this method data collection for both methods occur simultaneously and are analyzed separately. This aims to have two separate sets of data that help inform each other

2. Embedded Design - Here, both data sets are collected simultaneously, but the qual data is embedded within the quant data. This method is good when your focus in on the quant data but you need additional qual insights to explain it

3. Explanatory Sequential - This design involves collecting quant data first, followed by qual. It helps when you want to better explain your quant data set with additional qual insights

4. Exploratory Sequential - Here the researcher gathers qual data first, then quant data. This method helps when the researcher needs to explore a topic before backing it up with quant data.

How to Implement Mixed Methods in User Research

1. Define Learning Objective - Before you begin, define what it is you hope to learn then base the research methods around that learning objective

2. Choose Research Techniques - Select which qualitative and quantitative data collection techniques that will best help you meet your learning objectives. 

3. Conduct Research - Implement the chosen research methods, ensuring that they are aligned with your objectives. Use qualitative data to understand user motivations and behaviors and quantitative data to measure preferences.

4. Analyze Data Separately - Code qualitative data to look for themes and patterns and analyze quantitative data to look for statistical significance 

5. Compare and Triangulate Data - Compare both data sets to see where they converge or diverge from each other. Validate your insights using triangulation

6. Communicate Results to Stakeholders - Use strong presentation and storytelling skills to paint the fullest picture of your users to help sell stakeholders on your findings and the best path to move forward. 

When trying to understand user behavior, needs and preferences try combining qualitative and quantitative methods to effectively triangulate data. This use of mixed methods will lead to more comprehensive and reliable insights. This will help validate your research to stakeholders in leadership, product and development.