Statistics > Applications
[Submitted on 31 Jan 2015 (v1), last revised 10 May 2015 (this version, v2)]
Title:Quantifying replicability in systematic reviews: the r-value
View PDFAbstract:In order to assess the effect of a health care intervention, it is useful to look at an ensemble of relevant studies. The Cochrane Collaboration's admirable goal is to provide systematic reviews of all relevant clinical studies, in order to establish whether or not there is a conclusive evidence about a specific intervention. This is done mainly by conducting a meta-analysis: a statistical synthesis of results from a series of systematically collected studies. Health practitioners often interpret a significant meta-analysis summary effect as a statement that the treatment effect is consistent across a series of studies. However, the meta-analysis significance may be driven by an effect in only one of the studies. Indeed, in an analysis of two domains of Cochrane reviews we show that in a non-negligible fraction of reviews, the removal of a single study from the meta-analysis of primary endpoints makes the conclusion non-significant. Therefore, reporting the evidence towards replicability of the effect across studies in addition to the significant meta-analysis summary effect will provide credibility to the interpretation that the effect was replicated across studies. We suggest an objective, easily computed quantity, we term the r-value, that quantifies the extent of this reliance on single studies. We suggest adding the r-values to the main results and to the forest plots of systematic reviews.
Submission history
From: Liat Shenhav [view email][v1] Sat, 31 Jan 2015 09:54:42 UTC (468 KB)
[v2] Sun, 10 May 2015 13:28:10 UTC (619 KB)
References & Citations
Bibliographic and Citation Tools
Bibliographic Explorer (What is the Explorer?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)
Code, Data and Media Associated with this Article
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)
Demos
Recommenders and Search Tools
Influence Flower (What are Influence Flowers?)
Connected Papers (What is Connected Papers?)
CORE Recommender (What is CORE?)
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.