The multiversal methodology as a remedy of the replication crisis

Giulio Giacomo Cantone

No kuhmz, MetaArXiv from Center for Open Science

Abstract: This manuscript is a comprehensive historical and theoretical examination of the development of ‘multiversal methods’ as a response to the replication crisis. Multiversal methods are statistical procedures designed to assess the uncertainty arising from analyst-driven decisions in inferential models based on statistical regressions. The replication crisis is a surge in discovering that many studies fail to replicate the findings of previous studies. Replication crisis has raised concerns about the reliability and credibility of scientific research, particularly in social sciences and medicine. Section I provides a non-technical overview of the design of causal inference based on statistical regressions. Furtherly, it outlines and comments on the procedures to compute multiversal statistics. Section II presents the historical and social context within occurred key epistemological innovations contributing to the development of the theories behind multiversal methods. The section argues why and what these advancements drew from the epistemology of misinformation (‘bullshit epistemology’) for a sense of urgency for remedies to some enduring issues in scientific production: publication bias and p-hacking. Section III is a comment over two relevant works within paradigm of Open Science, to outline the limitations and challenges of this framework.

Date: 2023-04-26
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
https://osf.io/download/6449c9093848536ac6495ab3/

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:osf:metaar:kuhmz

DOI: 10.31219/osf.io/kuhmz

Access Statistics for this paper

More papers in MetaArXiv from Center for Open Science
Bibliographic data for series maintained by OSF ().