Multiverse analysis is a statistical approach in which researchers present all the results deriving from all combinations of arbitrary, but plausible decisions involved in every data analytic process (instead of focusing on a single result). Interpreting the multiverse of results provides fundamental insights into the robustness and uncertainty of scientific findings, therefore increasing transparency. Meta-analysis could particularly benefit from a multiverse approach, as it involves several decisions (e.g., how to impute missing data, which model to use, and how to deal with outliers). However, no specific framework is available to properly interpret the multiverse of meta-analytic results. Based on a real case study, the current work aims to propose an exploratory framework to support researchers in evaluating multiverse meta-analysis via tabular and graphical representations. Furthermore, we will highlight the contribution of each arbitrary decision to the variability of results.
Multiverse analysis is a statistical approach in which researchers present all the results deriving from all combinations of arbitrary, but plausible decisions involved in every data analytic process (instead of focusing on a single result). Interpreting the multiverse of results provides fundamental insights into the robustness and uncertainty of scientific findings, therefore increasing transparency. Meta-analysis could particularly benefit from a multiverse approach, as it involves several decisions (e.g., how to impute missing data, which model to use, and how to deal with outliers). However, no specific framework is available to properly interpret the multiverse of meta-analytic results. Based on a real case study, the current work aims to propose an exploratory framework to support researchers in evaluating multiverse meta-analysis via tabular and graphical representations. Furthermore, we will highlight the contribution of each arbitrary decision to the variability of results.
Multiverse Meta-Analysis: Proposing an Exploratory Framework
RASTI, SAJEDEH
2022/2023
Abstract
Multiverse analysis is a statistical approach in which researchers present all the results deriving from all combinations of arbitrary, but plausible decisions involved in every data analytic process (instead of focusing on a single result). Interpreting the multiverse of results provides fundamental insights into the robustness and uncertainty of scientific findings, therefore increasing transparency. Meta-analysis could particularly benefit from a multiverse approach, as it involves several decisions (e.g., how to impute missing data, which model to use, and how to deal with outliers). However, no specific framework is available to properly interpret the multiverse of meta-analytic results. Based on a real case study, the current work aims to propose an exploratory framework to support researchers in evaluating multiverse meta-analysis via tabular and graphical representations. Furthermore, we will highlight the contribution of each arbitrary decision to the variability of results.File | Dimensione | Formato | |
---|---|---|---|
Master Thesis.pdf
accesso aperto
Dimensione
1.58 MB
Formato
Adobe PDF
|
1.58 MB | Adobe PDF | Visualizza/Apri |
The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License
https://hdl.handle.net/20.500.12608/48646