In our series Interview With A Team Member, we aim to introduce the people behind the JASP project. Today we are interviewing Simon Kucharsky.
What is your professional background?
I studied Psychology at Charles University in Prague before doing my masters here at University of Amsterdam. I specialised in psychological methods and statistics. Now, I am starting a PhD under E.-J. Wagenmakers.
What is your favorite statistical test?
Even though these are not tests per se, I like anything with the word “latent”. In particular, I like procedures which classify things into latent groups, and so I grew in love with Latent Class Analysis, Latent Profile Analysis, and mixture models in general (and for that, I would like to thank Peter Edelsbrunner).
What is your relation to JASP?
I have worked for JASP since last year. My main task is to write a book that will accompany the recently released Data Library (a collection of more than 50 data examples that can be accessed directly in JASP). Other than that, I also helped with adding some small features in JASP, but the book writing part is my main priority.
What feature of JASP do you like best?
I am going to repeat what already others have said: I like that everything you do in JASP updates instantly in a nicely formatted output. This makes it so comfortable and fast! Although I generally prefer R, I use JASP a lot – whereas it would take me some time to do the analysis, create nice plots, format the output to proper tables and comment the code in R, I can do everything with JASP at once like snapping my fingers. The new filtering functionality is also very handy and makes a lot of things simple to do.
What aspect of JASP would you like to see improved in a future version?
Finishing the book :). Other than that, JASP is getting very close to be a comprehensive statistical package. We did comparisons with some statistical textbooks and JASP is not that far away from doing everything the books recommend (a huge shout-out to the development team for being so fast, and especially Jan Voelkel for making sure JASP can do whatever Andy Field thinks is good to know). But having multilevel models would be fantastic and I miss this feature the most. And of course, I am really looking forward to the day I open a JASP version in which I can edit the plots.
Are you a Bayesian, a frequentist, an agnostic, a pragmatist, or perhaps something else?
I specifically went to Amsterdam because of the department’s reputation for Bayesian stats. I prefer Bayesian statistics, although I have to admit I still have a high p-value/BF ratio. So if I define myself by how I think, I would say Bayesian, if I define myself by what have I done, I would have to say frequentist. But I would like to learn enough to afford being a pragmatist, i.e., using the appropriate framework depending on the data and questions we want to answer (so mostly Bayesian of course :)).
What question would you like to answer?
What kind of a Bayesian would you then be?
I don’t know. Sometimes I wonder where would I end up in I. J. Good’s “46656 Varieties of Bayesians”, but I change my opinions too quickly to categorise myself. Anyhow, it is a fun read.