If you analyse data, you need to communicate as well as calculate. I make data visualisations ready for publication, and interactive graphics and maps ready for the web.
The way statistics is taught is changing. I teach with the modern approach to an introductory course, bringing data science, machine learning and statistics together and emphasising computing, simulation and problem-solving with real data.
I have worked for 20 years on observational studies, clinical trials, mixed-methods research, clinical audit, systematic reviews, meta-analysis, scale validation and service evaluation. Now, I consult on design, analysis and communication of studies like these.
This website sets out my approach to animated charts with examples and code for R and Stata. I prefer to code from basics and retain full control, rather than take an easier approach and choose from software presets.
Writing and communicating
All that number-crunching is worthless if the message doesn't get across. I have twenty years of experience writing for audiences from technical experts to the general public, and from tweets to books. I can take analytic results and help turn them into a package that informs and achieves impact.
Bayesian methods give us intuitive measures for complex statistical problems. Latent variable models were a focus of my time in academia. New software (I am one of the developers of Stan) and parallel processing can provide answers in a fraction of the time.
There has never been a more exciting time to work in data science. My background is statistics (first degree mathematics) but I find maching learning techniques fascinating and have applied cutting-edge algorithms like deep learning and random forests. I help experts in their own fields to navigate the many options, avoid being ripped off, and interpret results carefully.