Introduction to statistical modelling: linear regression.

Research output: Contribution to journalArticlepeer-review

Abstract

In many studies we wish to assess how a range of variables are associated with a particular outcome and also determine the strength of such relationships so that we can begin to understand how these factors relate to each other at a population level. Ultimately, we may also be interested in predicting the outcome from a series of predictive factors available at, say, a routine clinic visit. In a recent article in Rheumatology, Desai et al. did precisely that when they studied the prediction of hip and spine BMD from hand BMD and various demographic, lifestyle, disease and therapy variables in patients with RA. This article aims to introduce the statistical methodology that can be used in such a situation and explain the meaning of some of the terms employed. It will also outline some common pitfalls encountered when performing such analyses.
Original languageEnglish
JournalRheumatology (Oxford, England)
Volume[Epub ahead of print]
DOIs
Publication statusPublished - 16 Apr 2013

Fingerprint

Dive into the research topics of 'Introduction to statistical modelling: linear regression.'. Together they form a unique fingerprint.

Cite this