Overview
Regression analysis is a very powerful technique that allows you to investigate the combined associations between one or more predictors and an outcome.
Some examples where this is helpful are:
- where within a trial you may wish to adjust for factors that differ between treatment groups to gauge the true effect of treatment
- in observational studies where you might want to take into account differences between the demographics or health behaviours of two or more subgroups
- when considering the combined effects of different factors, which may facilitate understanding of variation in outcome
Regression is a vital tool for any quantitative researcher.
This course takes you from the basics of types of regression to the formulation of a multiple linear regression model. Interaction terms are introduced and explained.
Concessions
A 50% discount is available for UCL staff, students, alumni. If you're eligible for a discount, email ich.statscou@ucl.ac.uk before booking to be sent the discount code.
The course is available for free to those associated with the Institute of Child Health or Great Ormond Street Hospital, and UCL doctoral students. Please also email ich.statscou@ucl.ac.uk to gain a booking code.
Learning outcomes
By the end of this course, participants will have an overview of different regression types and have the tools to perform and interpret the results from one specific type named 'linear regression'. The main course focuses on the theory behind regression analysis, and covers the formulation, interpretation and validation of linear regression models. In particular, participants will be able to:
- detail the basic principles behind regression analysis
- identify which type of regression is appropriate for a given research question and context
- perform a simple linear regression analysis, and be able to display graphically and interpret the output
- explain the merits of multiple regression and perform such an analysis
- include and interpret interaction terms in a model
- assess the assumptions of linear regression and interpret whether the fitted model is valid
- decide on a final model by considering potential variables for inclusion
Find out about other statistics courses
CASC's stats courses are suitable for anyone requiring an understanding of research methodology and statistical analyses. The courses allow non-statisticians to interpret published research and/or undertake their own research studies.
Find out more about CASC's full range of statistics courses.
