Five Tips to Get the Most Value from Human Factors Research: Leveraging Biometrics to Bridge the Gap Between Data and Empathy
On a recent project for one of the world’s largest tech firms, we conducted human factors research that generated data from eye tracking, galvanic skin responses, and facial expressions. Combined with qualitative responses, this research enabled our client to better understand exactly what users were thinking when interacting with a variety of screens, content, and calls to action. But beyond the insights uncovered for the client, we also learned of five ways to improve upon our next human factors research activities.
1. Create Well-Defined, Complex Tasks That Mimic an Actual User Experience
To provide deep insights into user frustrations and interests, scenarios need to be clearly articulated and tasks should be contextually relevant. When users engage in tasks that closely mirror real-life experiences, the data collected is more accurate and actionable. While we collaborated with the client to define the tasks, iterated on them until they were clearly defined, and then created a script to walk users through each activity, we learned that a best practice is to continuously refine the tasks throughout the study to ensure they remain relevant and challenging. If tasks turn out to be too simplistic, there may not be adequate opportunities for a participant to experience enough frustration or satisfaction to demonstrably show up in the biometric data.
2. Invest in Building Detailed Prototypes
For the study, our client built a series of prototypes. However, due to various constraints, they were limited in scope (e.g., participants could only interact with certain areas of the screen, which limited their ability to explore naturally). While the prototypes did let us gather eye-tracking data and qualitative feedback, they fell short of eliciting interesting biometric data. Investing in more detailed and interactive prototypes will provide a richer context for users, leading to more accurate and useful biometric readings. This data will then help identify usability issues and emotional responses that might not be apparent with more limited prototypes, which in turn can lead us to more user-centric designs.
3. Combine Biometric and Qualitative Insights for a Holistic View
Biometric data can be complex and ambiguous and requires nuanced interpretation. On its own, it may not provide the complete picture. Combining biometric data with qualitative inputs can help researchers more deeply understand a participant’s motivation and frustration. In addition, this hybrid approach can help get others on board with the findings and recommendations. Stakeholders can see quotes and video clips from real users while also seeing charts showcasing their eye movements and skin responses. It can sometimes be easy to dismiss one or the other, but it's hard to ignore both qualitative and human factors together.
4. Ensure Consistency in Testing Conditions
Consistency is not always required for qualitative projects, but in human factors research that involve biometric readings, it is essential. Whenever possible, we suggest ensuring that pages or screens being compared are similar in content and functionality to avoid skewed results. When necessary, aggregate metrics to provide a holistic view of user interactions, ensuring that your comparisons are fair and accurate. In addition, we recommend a standardized protocol, such as a script and uniform instructions for participants, as well as common environmental conditions to minimize external influences on the data. For this project, our team used a dedicated research facility to create and manage ideal conditions for all participants.
5. Use Data Visualization to Highlight Key Findings but Show Where the Data May Not Show Value
The amount of data generated by eye tracking and skin response, for example, is huge. It’s very easy to overlook something important in tables or text, so data visualization is a necessity. Tables and the full dataset should always be made available, but it’s unlikely many will dig that deep. As mentioned before, finding insight often involves nuanced interpretation, so we highly recommend being transparent about the research, including what data is useful and what is not. For example, in this project, we suggested that while we collected a large amount of biometric data, much of it wasn’t as useful as the eye tracking because of limitations of the scenarios and prototypes. Transparency about what was useful and what was noise goes a long way with getting stakeholders to understand the full scope of the findings and encourage further research.
Conclusion
Studying human behavior is essential to designing products and systems that meet real user needs, and human factors research helps organizations better understand exactly how users interact with specific products. We believe the five insights we have outlined can make everyone’s human factors research more effective and impactful.
Keep in Touch
Our team has helped hundreds of industry leaders across SaaS, life sciences and other industries to solve many of the challenges we’ve discussed here. If you’d like to discuss ways our team can help yours with UX research, strategy and product design to move your products forward, please reach out at info@proximitylab.com.
On a recent project for one of the world’s largest tech firms, we conducted human factors research that generated data from eye tracking, galvanic skin responses, and facial expressions. Combined with qualitative responses, this research enabled our client to better understand exactly what users were thinking when interacting with a variety of screens, content, and calls to action. But beyond the insights uncovered for the client, we also learned of five ways to improve upon our next human factors research activities.
1. Create Well-Defined, Complex Tasks That Mimic an Actual User Experience
To provide deep insights into user frustrations and interests, scenarios need to be clearly articulated and tasks should be contextually relevant. When users engage in tasks that closely mirror real-life experiences, the data collected is more accurate and actionable. While we collaborated with the client to define the tasks, iterated on them until they were clearly defined, and then created a script to walk users through each activity, we learned that a best practice is to continuously refine the tasks throughout the study to ensure they remain relevant and challenging. If tasks turn out to be too simplistic, there may not be adequate opportunities for a participant to experience enough frustration or satisfaction to demonstrably show up in the biometric data.
2. Invest in Building Detailed Prototypes
For the study, our client built a series of prototypes. However, due to various constraints, they were limited in scope (e.g., participants could only interact with certain areas of the screen, which limited their ability to explore naturally). While the prototypes did let us gather eye-tracking data and qualitative feedback, they fell short of eliciting interesting biometric data. Investing in more detailed and interactive prototypes will provide a richer context for users, leading to more accurate and useful biometric readings. This data will then help identify usability issues and emotional responses that might not be apparent with more limited prototypes, which in turn can lead us to more user-centric designs.
3. Combine Biometric and Qualitative Insights for a Holistic View
Biometric data can be complex and ambiguous and requires nuanced interpretation. On its own, it may not provide the complete picture. Combining biometric data with qualitative inputs can help researchers more deeply understand a participant’s motivation and frustration. In addition, this hybrid approach can help get others on board with the findings and recommendations. Stakeholders can see quotes and video clips from real users while also seeing charts showcasing their eye movements and skin responses. It can sometimes be easy to dismiss one or the other, but it's hard to ignore both qualitative and human factors together.
4. Ensure Consistency in Testing Conditions
Consistency is not always required for qualitative projects, but in human factors research that involve biometric readings, it is essential. Whenever possible, we suggest ensuring that pages or screens being compared are similar in content and functionality to avoid skewed results. When necessary, aggregate metrics to provide a holistic view of user interactions, ensuring that your comparisons are fair and accurate. In addition, we recommend a standardized protocol, such as a script and uniform instructions for participants, as well as common environmental conditions to minimize external influences on the data. For this project, our team used a dedicated research facility to create and manage ideal conditions for all participants.
5. Use Data Visualization to Highlight Key Findings but Show Where the Data May Not Show Value
The amount of data generated by eye tracking and skin response, for example, is huge. It’s very easy to overlook something important in tables or text, so data visualization is a necessity. Tables and the full dataset should always be made available, but it’s unlikely many will dig that deep. As mentioned before, finding insight often involves nuanced interpretation, so we highly recommend being transparent about the research, including what data is useful and what is not. For example, in this project, we suggested that while we collected a large amount of biometric data, much of it wasn’t as useful as the eye tracking because of limitations of the scenarios and prototypes. Transparency about what was useful and what was noise goes a long way with getting stakeholders to understand the full scope of the findings and encourage further research.
Conclusion
Studying human behavior is essential to designing products and systems that meet real user needs, and human factors research helps organizations better understand exactly how users interact with specific products. We believe the five insights we have outlined can make everyone’s human factors research more effective and impactful.
Keep in Touch
Our team has helped hundreds of industry leaders across SaaS, life sciences and other industries to solve many of the challenges we’ve discussed here. If you’d like to discuss ways our team can help yours with UX research, strategy and product design to move your products forward, please reach out at info@proximitylab.com.
