Galbraith plot

In statistics, a Galbraith plot (also known as Galbraith's radial plot or just radial plot), is one way of displaying several estimates of the same quantity that have different standard errors.[1]

Example for Galbraith's radial plot.

It can be used to examine heterogeneity in a meta-analysis, as an alternative or supplement to a forest plot.

A Galbraith plot is produced by first calculating the standardized estimates or z-statistics by dividing each estimate by its standard error (SE). The Galbraith plot is then a scatter plot of each z-statistic (vertical axis) against 1/SE (horizontal axis). Larger studies (with smaller SE and larger 1/SE) will be observed to aggregate away from the origin.[2]

See also

References

  1. Galbraith, Rex (1988). "Graphical display of estimates having differing standard errors". Technometrics. Technometrics, Vol. 30, No. 3. 30 (3): 271–281. doi:10.2307/1270081. JSTOR 1270081.
  2. University of York, Department of Health sciences MSc course material

External links

Further reading

This article is issued from Wikipedia - version of the 8/10/2016. The text is available under the Creative Commons Attribution/Share Alike but additional terms may apply for the media files.