| Literature DB >> 26061676 |
Mark Crowther1, Alison Avenell2, Graeme MacLennan1, Graham Mowatt1.
Abstract
When performing a systematic review, whether or not a meta-analysis is performed, graphical displays can be useful. Data do still need to be described, ideally in graphical form. The Harvest plot has been developed to display combined data from several studies that allows demonstration of not only effect but also study quality. We describe a modification to the Harvest plot that allows the presentation of data that normally could not be included in a forest plot meta-analysis and allows extra information to be displayed. Using specific examples, we describe how the arrangement of studies, height of the bars and additional information can be used to enhance the plot. This is an important development, which by fulfilling Tufte's nine requirements for graphical presentation, allows researchers to display evidence in a flexible way. This means readers can follow an argument in a clear and efficient manner without the need for large volumes of descriptive text.Keywords: Harvest plot; data interpretation; graphical representation; systematic review
Year: 2011 PMID: 26061676 DOI: 10.1002/jrsm.37
Source DB: PubMed Journal: Res Synth Methods ISSN: 1759-2879 Impact factor: 5.273