How I fell in love with R

The fun part of doing a PhD are the PhD courses. I didn’t realize this at the beginning, when they informed me that Vrije Universiteit obligates PhD candidates to obtain 30 credits. But by now (I have obtained 25 credits in 2 years time) I am very happy with this obligation. I have learned so much which I can not only use in my PhD project but also as a lecturer.

My favorite courses are those about R. R is a language and environment for statistical computing and graphics. R is available as free software under the terms of the Free Software Foundation’s GNU General Public License in source code form. It compiles and runs on a wide variety of UNIX platforms and similar systems (including FreeBSD and Linux), Windows and MacOS. My first statistical course using R was in April 2016 at the Vrije Universiteit, and I was pretty nervous. I haven’t done many courses on statistics and whenever my computer does something I don’t want it to do I panic. The same happened during this course: I totally panicked. I didn’t understand the software, and I also ran into trouble with the statistical analysis I was supposed to do in R. It wasn’t the best moment in my life. My children were aged 4 and 5 and I had worked way too much and way too hard for over a year. During the course I realized I was not able to do the exam, I needed to take a break.

So I did and followed the same course one year later. Still I had difficulty understanding the way R works but I managed to copy the syntaxes from my lecturer and I was able to interpret the output. I passed the exam and put all I had learned about R in a folder and stored it. In October 2017 I was ready to analyze my first data, so I opened R and not SPSS. And that was the moment I fell in love with R. I managed to do all the things I needed to do, and all the things I didn’t know I could find on the internet. When multilevel modeling appeared not to be possible (my number of respondents per country was too low) I changed to structural equation modeling comparing groups. A new challenge which I managed to master more or less. Enough to get a submission to a congress I really wanted to go to accepted (EMAC Doctoral Colloquium 2018)

Last week I did a course on Strutural Equation Modeling using Lavaan in R by prof Rosseel, the developer of the lavaan package. And this time I could fully understand what he was explaining, even the algebra matrix which is the basis for the algorithms in the package. In two weeks’ time I have to give a course using SPSS. That will be challenging as I hardly ever use the buttons anymore. Even in SPSS I prefer writing syntaxes (although that is frustrating to do as this software is not build for that). My aim for the near future is to teach statistics with R and it seems to be heading in the right direction. In the meantime I have lots of data to analyze in R!