Richard specializes in methodology and statistics, and has made significant contributions to the area of Bayesian statistics. Anyone who is computing Bayes factors in R is likely using the ‘BayesFactor’ package
of which Richard is the first author (together with Jeff Rouder
and Tahira Jamil). The last digital event on this website that dedicated three posts to confidence intervals (here
, and here
) was also stimulated by Richard’s work.
It will therefore come as little surprise that Richard’s bachelor's degree is …err… in music, with a minor in philosophy.
As Richard puts it: “I was fascinated by music theory, and did quite a lot of research on tuning systems. I enjoy performing, but the method of music held a fascination for me that it didn't for other singers I knew. When I started as a PhD student in psychology working with Jeff Rouder, I took my first statistics class, and knew from then on I've been interested in methods. What unites all my work -- both the modelling side and the general inference side -- is an interest in the logic of learning from data.”
Richard has managed not only to combine music and Bayesian statistics, but he also managed to attend the Psychonomics conference since before he became a psychologist. In 2000, while he was finishing up his bachelor's degree in music, Richard attended the Psychonomics meeting in New Orleans with Candice Morey, his fiancée at the time who was wrapping up a bachelor's degree in Psychology. Candice is now at the University of Edinburgh
in Scotland and Richard is at Cardiff University in Wales. (An easy commute, all that separates them is England).
Richard is also a vocal champion of “open data” and “open science”: All his software is open source and he believes—like many others—that psychology should move towards a more sharing-oriented culture, with code, data, and materials being publicly available. I will post on open data and open science next week, so stand by for more news about several of those initiatives.
Finally, Richard’s appointment as Digital Associate Editor comes with a challenge to the membership: He is the world’s best judge of line lengths—or, at least the best judge whose performance has been confirmed experimentally. This was documented in 2004 in an article in Psychonomic Bulletin & Review
. (Subject confidentiality is not an issue for participants who are authors and set world records.)
George Miller famously suggested that "seven plus or minus two" was the limit for a line identification task; however, Richard reached ceiling on 20 stimuli. This means that for the lines in the figure below, he was able to assign the correct number (shown on the left) to each stimulus when it was presented in isolation with 100% accuracy:
Anyone wishing to challenge this record should submit their results to Psychonomic Bulletin & Review.