Because many of my research questions in psychology are not amenable to
experimental designs, I have become interested in statistics and computational
statistics. I particularly focus on structural equation models, mixed
effects models, and data visualization. My own work is almost exclusively
done in R; however, when I am collaborating with others I am comfortable working
with Mplus, EQS, or SAS.
In 2011, I joined the UCLA Institute for Digital Research and Education
Statistical Consulting Group. One aspect I enjoy as a consultant is having
the opportunity to hear about and assist with research from departments all over
campus. In addition to my work as a consultant, I have tutored graduate
psychology statistics at UCLA, I am an active member of the R
responses on the R-help listserv, and answer questions on
This is some of the feedback I have received from clients I have worked with.
The packages included EQS, Mplus, R, SAS, and Stata with problems anywhere from
data management, power analysis, and regression models to complex mixed effects
models and latent variable models. For full disclosure, I do get some
lower ratings (6s and 7s) but comments are optional and less happy clients never
seem to leave any feedback.
- Josh was extremely helpful in explaining the model and and output. He's
great asset to ATS --- 10/10 May 2012
- Just wanted to verify some quick things as I'm getting ready to submit a
manuscript. Josh was very helpful as always! --- 10/10 April 2012
- Thanks Josh for help with using G Power calculations for upcoming grant
submission to NIH to study biomarkers of illness progression in Huntington's
Disease. --- 10/10 April 2012
- Josh is awesome and helped me with problems in both Stata and MPlus.
--- 10/10 April 2012
- Josh was really fantastic! He walked me through multilevel mediation quickly
and clearly, and was extremely helpful. He also showed me some helpful tips in
SAS that will make my life much easier. --- 10/10 March 2012
- Josh helped me run mixed-effects negative binomial models in R using package
"glmmADMB" --- 10/10 January 2012
- Josh helped me run mixed-effects Cox models on some behavioral data on
lizards (using R package "coxme") --- 10/10 February 2012
- Josh is always very knowledgeable and extremely helpful --- 10/10
- Thanks very much to Josh, who taught me new things today! --- 10/10
- Josh was extremely helpful and helped me solve my issues. --- 10/10
- I wish that I had utilized these services beforehand. Josh was very helpful.
- A series of comments from one particularly enthusiastic client:
All praises to Josh!
Thanks once again to R wizard Josh.
Josh is a genius.
Josh did a phenomenal job helping me to prepare my dataset.
Right now I do not have any particular projects going on.
I have taught a number of mini-courses on introductory statistics to small
I have also led several one session seminars using R. As I do more of
these and create more resources, I am hoping to put together cohesive sets
(lectures, examples, exercises, etc.) on specific topics.
On the far distant horizon, I am hoping to use these as chapters in a free (as
in speech and as in beer) book on introductory
statistics for the social sciences. This would be roughly modelled after
various open source software projects, and ideally would be something of a collaborative/community project.