Note: This iteration of the course is a special event open only to Wellcome Sanger Institute students and staff. If you are a Sanger staff member, please register your interest here to be notified once a place becomes available. After accepting an offer, a place will be booked and you will receive a confirmation email with relevant payment details.

Programme

The Sanger Institute Graduate Computational Skills Programme

Module 7: Extending the linear model

During the Core statistics sessions we provided a strong foundation in practical statistics and data analysis. We highlighted that many statistical tests are actually all variations on the same principle: the linear model.

The linear model is appropriate in many cases, where the response variable is continuous and a set of underlying statistical assumptions are met. However, what to do when this is not the case?

In this one-day course we give a brief overview of what your options are when the standard linear model is no longer suitable. We cover situations where your response variable is non-continuous (generalised linear models) or your data are not independent (linear mixed models).

We discuss this in the context of experimental design: having awareness of what your data will look like by the end of the experiment can avoid common pitfalls, such as treating technical repeats as biological ones or treating count data as continuous.

This module consists of 1 full day sessions.

Delivery

All sessions are in-person.