XClose

Short courses

Home
Menu

Using Simulation Studies to Evaluate Statistical Methods

  • 10 hours
  • 2 days

Overview

Simulation studies are an important tool for statistical research. They help statisticians and researchers understand the properties of statistical methods and compare different methods.

This two-day course will help you understand how simulation studies work, so you can critique published simulation studies and design one yourself.

You'll learn how to plan, code, analyse and interpret simulation studies using Stata or R (data analysis and statistical software). 

This course is run by the Institute of Clinical Trials and Methodology at UCL.

Course content and structure

Topics covered by the course include the following:

  • Planning a simulation study
  • Coding a simulation study
  • Analysing simulation studies
  • Analysing your simulation study and feeding back results
  • Reporting simulation studies

The course involves a series of lectures and computer practicals, where you can use Stata or R.

Examples will be taken from the lecturers' experiences, primarily relating to medical statistics. The principles for simulation studies in other applied areas are the same, but the examples may be less relevant.

Who this course is for

This course is suitable for:

  • methodological or applied statisticians who need to evaluate the statistical properties of one or more methods
  • PhD students who use simulation studies
  • readers of methodology articles that evaluate methods by simulation

Entry requirements

You should already be familiar with Stata or R and know, for example, how to generate data, run a regression command and produce simple graphs.

You'll need to bring your own laptop with Stata (version 12 or later) or a recent version of R installed.

Learning outcomes

By the end of this course you'll be able to:

  • critically read and evaluate simulation studies in the statistical literature
  • conduct a simulation study
  • explain the rationale for simulation
  • understand the importance of careful planning
  • code and debug simple simulation studies in Stata
  • analyse simulation studies producing estimates of uncertainty
  • present methods and results for publication

Cost and concessions

The fees are:

  • standard - £270
  • UCL staff and students - £150
  • staff and students based at the ICTM - free

Course team

Dr Tim Morris

Dr Tim Morris

Tim is a statistician interested in practical methods for improving the design and analysis of randomised trials and observational studies. He's based at the MRC Clinical Trials Unit at UCL, part of the Institute of Clinical Trials and Methodology.

Professor Ian White

Professor Ian White

Ian is a medical statistician with an interest in developing new methodology for design and analysis of clinical trials, meta-analysis and observational studies. He joined the MRC Clinical Trials Unit at UCL in 2017 after spending 16 years as a programme leader at the Medical Research Council's Biostatistics Unit in Cambridge.

Dr Michael J. Crowther

Dr Michael J. Crowther

Michael is Associate Professor of Biostatistics in the Biostatistics Research Group at the University of Leicester. He's a Section Editor for the Journal of Statistical Software and an Associate Editor for the Stata Journal. His main research interests include survival analysis, multilevel and mixed effects models, and statistical software development.

Learner reviews

"Despite having run simulation studies in the past I learned so much. Thank you."

"Tutors were really helpful during practicals. It was good that we had the opportunity to think about and develop simulations from scratch with cheat-sheets available if required."

Course information last modified: 30 Nov 2022, 16:00