Choosing effect measures and computing estimates of effect

Julian P.T. Higgins, Tianjing Li, Jonathan J. Deeks

Research output: Chapter in Book/Report/Conference proceedingChapter

Abstract

A key early step in analysing results of studies of effectiveness is identifying the data type for the outcome measurements. This chapter considers outcome data of five common types: dichotomous (or binary) data, continuous data, ordinal data, count or rate data and time-to-event data. The ways in which the effect of an intervention can be assessed depend on the nature of the data being collected. For each of the types of data, the chapter reviews definitions, properties and interpretation of standard measures of intervention effect, and provides tips on how effect estimates may be computed from data likely to be reported in sources such as journal articles. Formulae to estimate effects for the commonly used effect measures are provided in a supplementary document statistical algorithms in Review Manager, as well as other standard textbooks. Effect measures are either ratio measures or difference measures. Ratio measures are typically analysed on a logarithmic scale.

Original languageEnglish (US)
Title of host publicationCochrane Handbook for Systematic Reviews of Interventions
Publisherwiley
Pages143-176
Number of pages34
ISBN (Electronic)9781119536604
ISBN (Print)9781119536628
DOIs
StatePublished - Jan 1 2019

Keywords

  • Computing estimates
  • Continuous data
  • Count data
  • Dichotomous data
  • Difference measures
  • Effect measures
  • Ordinal data
  • Outcome measurements
  • Ratio measures
  • Time-to-event data

ASJC Scopus subject areas

  • Medicine(all)

Fingerprint Dive into the research topics of 'Choosing effect measures and computing estimates of effect'. Together they form a unique fingerprint.

Cite this