Another good class I took as part of the Big D Conference was presented by Eva Kaniasty, the founder of Red Pill UX, and a research and design consultancy.

The role of the UX researcher is an important one. We, as UX researchers, need to design our research studies for analysis. Obviously when we perform a story, we are trying to gather important data. This data we gain in our research efforts need to be analyzed and our findings need to be communicated to others. We need to think about how to visualize our research.

Get your stakeholders to empathize with their customers and users. One way to do this is to take photos of the real people using the product. Don’t use fancy stock photography with posed fake models. Use your smartphone and take pictures of people using the product. And take more pictures of the person, sort of posed, to use as your persona image. This makes the persona more realistic and will provide the opportunity for your stakeholders to see the real person behind the persona.

I learned about the website UI Faces where you can go and get more “realistic” photos that are free to use in your personas or other needs. Granted, I checked this site out, and there’s a lot of avatars from people I follow on Twitter. But hey, your customer probably does not follow them and therefore they won’t recognize the images. So go ahead and check out the site to see if it needs your image needs for personas.

The problem with personas today is that many people just make them up. They don’t generate them using interview data or base them on real users. People often create personas based on “ideal” customers which is not accurate. Be sure that when you create personas, create them based on real research. Also make sure that they represent real people and customers, not ideal ones.

Additional notes from this talk

• Pie charts are poor visualization tools much of the time.
• Icons can be used to visualize data, but don’t over use them.
• After you have a research session, write a quick summary right afterwards so you don’t forget the important details. The longer you wait, the more you will forget.
• Videos are time consuming and become outdated quickly.
• Quotes can be very powerful and easier to generate than video clips.
• Look for patterns in your data.
• Don’t use a word cloud to summarize data.
• Word clouds are hard to read, noisy and the colors used can be confusing, portraying a confusing hierarchy.
• A treemap shows the frequency of terms used in a combined bar chart.
• Make any color coding meaningful and explain what it means.
• Test with color blindness tools to make sure that color can be seen.
• Do no over aggregate that data. That happens when you smooth and combine data together too much. When this happens, the data can lose its meaning. Don’t combine much because if you do, you can lose where the problems are.
• Use words instead of illustrating with a bunch of repetitive icons.
• Don’t use statistics for something subjective like severity ratings.
• For “Ease of Use” ratings, use a bar chart, not a pie chart.
• Stars are not good to rate the severity of something. People think more stars means “good” and that is the opposite mental model for the severity rating scale.
• Dot voting is good to give everyone a chance to vote and it surfaces up the problems that need addressing first. The most votes wins!

Top visualization mistakes

• Implying statistical significance
• Over aggregation
• Comparing apples and oranges
• Leaving out context