Statistical Services Centre Ltd
  • Home
  • Training
    • Course Registration
    • Terms and Conditions
    • Bespoke Training
    • Mailing List Request
  • Consultancy
    • Expertise
  • Team
    • James Gallagher
    • Sandro Leidi
    • Cathy Garlick
  • SSC-Stat
  • Get Started
  • Contact us
  • Products
  • Blog

Introduction to Multiple Comparison Methods

Overview
The course consistes of a mixture of presentations and practicals.  It begins with an overview of the multiplicity problem, basic definitions and some important historial methods.  We then move onto parametric multiple inference in the linear model, and introduce the max T method which forms the basis for procedures such as Tukey and Dunnett's multiple comparison procedures.
The remainder of the course concentrates on more recent and increasingly popular multiple comparison methods, particularly in clinical trials.  The Closure Testing Principle, which forms the basis of many methods, is introduced and common stepwise procedures discussed.  We then move from "data driven" methods to fixed sequence methods and their variations.  Gatekeeping procedures for controlling type 1 error in the face of multiple objectives and the inclusion of primary, secondary and tertiary endpoints in regulatory claims are discussed.
The course finishes with further consideration on the analysis of multiple endpoints.

Who Should Attend?
Medical statisticians, and related, engaged in clinical trials and medical related research, who have little or no experience of dealing with multiplicity issues. Experience of SAS and R would be advantageous but is not essential.

How You Will Benefit
This course gives a practical introduction to multiple comparison methods from basic methods through to more recent sophisticated approaches.

What Do We Cover?
  • Basic concepts, simple p-value based methods: Bonferroni, Sidak, and Simes. Adjusted p-values and simultaneous confidence intervals.  Types of multiple comparison procedures
  • Parametric multiple comparison methodology in linear models: the max T method. Tukey and Dunnett's procedures as special cases. Use of Monte Carlo simulation
  • The Closed Testing Principle: stepwise methods
  • Sequential testing of pre-specified hypotheses: the fixed-sequence testing method, fall-back procedure. Gatekeeping methods
  • Further consideration of multiple endpoints, global procedures.

Software
Practical work will be done in: SAS supplemented by R where required.
Note: For practical work, participants must bring their own laptop with fully licensed versions of the software.



Course Dates
Next run to be announced

Duration: 1 day
Price: £TBC

Apply Now
(terms and conditions apply)

Return to full course listing
  • Home
  • Training
    • Course Registration
    • Terms and Conditions
    • Bespoke Training
    • Mailing List Request
  • Consultancy
    • Expertise
  • Team
    • James Gallagher
    • Sandro Leidi
    • Cathy Garlick
  • SSC-Stat
  • Get Started
  • Contact us
  • Products
  • Blog