Fast and Rigorous User Personas
Monday, March 14
11:00AM - 12:00PM
110 E 2nd St
There is a belief in the UX community that user personas can be either fast or rigorous, but not both. We present a fast and rigorous method based on cluster analysis of surveyed user traits (e.g. behaviors, attitudes). First, we build a survey to measure the space of user traits. Second, we submit the survey data to factor analysis to find groups of correlated traits. Third, we submit the factor scores to cluster analysis to identify groups of users who have similar traits. We demonstrate the method in the context of developing car shopper personas. We suggest best practices for building a meaningful survey, interpreting the factors, optimizing the clusters, and visualizing the results.
Data Scientist & Researcher
Currently a data scientist at Rackspace, I have over 10 years experience in data science and behavioral research. I earned my PhD in Cognitive Psychology, with a focus on learning and reasoning. I’...Show the rest
Currently a data scientist at Rackspace, I have over 10 years experience in data science and behavioral research. I earned my PhD in Cognitive Psychology, with a focus on learning and reasoning. I’m a “full stack” researcher, used to taking projects from conception, through stakeholder alignment, planning, execution, analysis, and sharing. My favorite part of the job is connecting big ideas to rigorous statistical models and distilling the results into a sound bite or a beautiful image. I crave opportunities to make things better and do things right.
On the technical end, I'm a huge advocate and 10 year veteran user of R, with some Python on the side. My strong suits are GLM, Bayesian inference, multivariate stats, and various machine learning techniques (e.g., decision trees, clustering, neural nets, NLP). Truth be told, I love data. Pretty much any data.Hide the rest
VP of Research
I am a User Experience (UX) Researcher, Strategist, Evangelist & Data Visualization addict. Currently Vice President of Research at CarStory.