A D-vine copula mixed model for joint meta-analysis and comparison of diagnostic tests
- Submitting institution
-
The University of East Anglia
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 182621745
- Type
- D - Journal article
- DOI
-
10.1177/0962280218796685
- Title of journal
- Statistical Methods in Medical Research
- Article number
- -
- First page
- 3286
- Volume
- 28
- Issue
- 10-11
- ISSN
- 0962-2802
- Open access status
- Compliant
- Month of publication
- October
- Year of publication
- 2019
- URL
-
https://arxiv.org/abs/1805.09674
- Supplementary information
-
-
- Request cross-referral to
- -
- Output has been delayed by COVID-19
- No
- COVID-19 affected output statement
- -
- Forensic science
- No
- Criminology
- No
- Interdisciplinary
- No
- Number of additional authors
-
0
- Research group(s)
-
-
- Citation count
- 2
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- Published in a highly ranked journal, proposes a novel and general model for joint meta-analysis and comparison of diagnostic tests that is not restricted to normality assumption, and in so doing not only unifies, but also improves on the recommended model in the literature, namely the quadrivariate generalized linear mixed model. Here instead of a quadrivariate normal distribution we employ a quadrivariate vine copula that provides general non-linear dependence and computational feasibility. In fact, we propose a numerically stable maximum likelihood estimation technique. Our algorithms are part of our open-source R package CopulaREMADA, which has over 26,000 downloads.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -