In this paper, we consider the meta-analysis approach as a special case ofmultilevel modelling in order to identify the potential sources of dissimilarities amongpre-election poll accuracy measures. The predictive accuracy measure is computed tocompare the pre-electoral poll result to the actual result. It is introduced in ahierarchical model as a dependent variable and the poll characteristics as explanatoryvariables. In the model each outcome is affected by a specific sampling error assumedto have a normal distribution and a known variance. Employing such a multilevelapproach, we conduct a meta-analysis of pre-election poll accuracy measures inParliamentary elections in Italy from 2001 to 2008.
|Titolo:||Multilevel models in meta-analysis of pre-election poll accuracy measures|
TOMASELLI, Venera (Corresponding)
|Data di pubblicazione:||2013|
|Citazione:||Multilevel models in meta-analysis of pre-election poll accuracy measures / D'Agata, ROSARIO GIUSEPPE; Tomaselli, Venera. - In: STATISTICA APPLICATA. - ISSN 1125-1964. - 23:2(2013), pp. 175-192.|
|Appare nelle tipologie:||1.1 Articolo in rivista|