Planning and Conducting Simulation Studies in R (Part 1)

Half day short course (9:30am – 12:30pm)

Short Course #3 

In the first part of this course, we will start with an introduction on how to write and structure your own R functions for conducting simulation studies, and will also hint at R packages providing templates for such functions. We will discuss important aspects for planning and conducting a simulation study, such as the choice between fully crossed and other experimental designs as well as strategies for quality control. We will further introduce typical outcome variables of simulation studies. For these, we will first show intuitively that the precision of the results depends on the number of iterations and how one can check informally whether the current number of iterations is sufficient. Then we will go into some more detail about the quantification of Monte Carlo standard errors and how they can be used for formally planning the number of iterations.

The course will consist of a mix of slide presentations and hands-on exercises in R. The participants bring their own laptops with R (and if they like RStudio) installed. During the exercises, the instructors will walk around and help the participants working on their own laptops.

Intended Audience

The intended audience of the first part of the course is undergraduate and graduate students, faculty and researchers who have used R before, but don’t have extensive experience with writing their own simulation studies yet, or would like to systematically review the steps and decisions necessary when planning a simulation study. For this audience it is possible to book only the first part of the course in the morning, or to also book the second part in the afternoon.

About the instructors

Carolin Strobl

 Carolin Strobl is a professor for Psychological Methods at the University of Zurich, Switzerland, with a background in psychology and statistics. Her group has contributed to several software packages related to machine learning and psychometrics in the free, open source software R. She has conducted various simulation studies over the years and has extensive expertise in teaching statistics and machine learning using R across various academic levels. This year, Carolin and her team will also publish a textbook on conducting simulation studies in R.

Mirka Henninger

 Mirka Henninger is an assistant professor for Statistics & Data Science at the University of Basel, Switzerland. Her background is in psychology, and her research is grounded at the intersection of psychometrics, machine learning, and multilevel modelling. Herein, she has conducted many simulation studies, and has been teaching statistics together R as a software tool for data analysis and programming at various academic levels. Mirka is also part of the team around Carolin publishing a textbook on how to plan and realize simulation studies in R.