A comparison of 20 heterogeneity variance estimators in statistical synthesis of results from studies

Πετροπούλου, Μαρία; Μαυρίδης, Δημήτρης

A comparison of 20 heterogeneity variance estimators in statistical synthesis of results from studies

dc.contributor.author	Πετροπούλου, Μαρία	el
dc.contributor.author	Μαυρίδης, Δημήτρης	el
dc.date.accessioned	2018-05-22T10:50:15Z
dc.date.available	2018-05-22T10:50:15Z
dc.identifier.uri	https://olympias.lib.uoi.gr/jspui/handle/123456789/29117
dc.rights	An error occurred on the license name.	*
dc.rights.uri	An error occurred getting the license - uri.	*
dc.subject	bias	en
dc.subject	type I error	en
dc.subject	heterogeneity variance estimators	en
dc.subject	power	en
dc.subject	simulation study	en
dc.title	A comparison of 20 heterogeneity variance estimators in statistical synthesis of results from studies	en
heal.abstract	When we synthesize research findings via meta-analysis, it is common to assume that the true underlying effect differs across studies. There is a plethora of estimation methods available for the between-study variability. The widely used DerSimonian and Laird estimation method has been challenged but knowledge for the overall performance of heterogeneity estimators is incomplete. We identified 20 heterogeneity estimators in the literature and evaluated their performance in terms of bias, type error I rate and power via a simulation study. Moreover, we compared the Knapp and Hartung and the Wald-type method for estimating confidence interval for the summary estimate. Although previous simulation studies have suggested the Paule-Mandel (PM) estimator, it has not been compared with all the available estimators. For dichotomous outcomes, estimating heterogeneity through Markov Chain Monte Carlo is a good choice if the prior distribution for heterogeneity is informed by published Cochrane reviews. Non parametric bootstrap (DLb) performs well for all assessment criteria for both dichotomous and continuous outcomes. The positive DerSimonian and Laird (DLp) and the Hartung-Makambi (HM) estimators can be an alternative choice for dichotomous outcome when the heterogeneity values are close to 𝟎.𝟎𝟕 and for continuous outcome for all and for medium heterogeneity values (𝟎.𝟎𝟏,𝟎.𝟎𝟓), respectively. Hence, they are heterogeneity estimators (DLb; DLp) which perform better than the suggested PM. Maximum likelihood (ML) provide the best performance for both types of outcome in the absence of heterogeneity.	en
heal.access	free
heal.bibliographicCitation	-	en
heal.dateAvailable	2018-05-22T10:51:15Z
heal.fullTextAvailability	true
heal.journalName	Statistics in Medicine	en
heal.journalType	peer-reviewed
heal.language	en
heal.publicationDate	2017
heal.publisher	Wiley	en
heal.recordProvider	Πανεπιστήμιο Ιωαννίνων. Σχολή Επιστημών Αγωγής. Παιδαγωγικό Τμήμα Δημοτικής Εκπαίδευσης.	el
heal.secondaryTitle	A simulation study	en
heal.type	journalArticle
heal.type.el	Άρθρο περιοδικού	el
heal.type.en	Journal article	en