Scripts for Post-Processing Mega-Meta Screening Results

by van den Brand, Sofie, Hofstee, Laura, Teijema, Jelle, Melnikov, Valentin, Kramer, Bianca, Brouwer, Marlies and Van de Schoot, Rens
Abstract:
The repository is part of the so-called, Mega-Meta study on reviewing factors contributing to substance use, anxiety, and depressive disorders. The study protocol has been pre-registered at Prospero. The procedure for obtaining the search terms, the exact search query, and selecting key papers by expert consensus can be found on the Open Science Framework. The three datasets, one for each disorder, used for screening in ASReview and the partly labeled output datasets can be found on DANS[NEEDS LINK]. The current repository contains the post-processing scripts to: (1) Merge the three output files after screening in ASReview; (2) Obtain missing DOIs; (3) Apply another round of de-duplication (the first round of de-duplication was applied before the screening started); (4) Deal with noisy labels corrected in two rounds of quality checks. The scripts in the current repository result in one single dataset that can be used for future meta-analyses. The dataset itself is available on DANS[NEEDS LINK].
Reference:
Scripts for Post-Processing Mega-Meta Screening Results (van den Brand, Sofie, Hofstee, Laura, Teijema, Jelle, Melnikov, Valentin, Kramer, Bianca, Brouwer, Marlies and Van de Schoot, Rens), Zenodo, 2021.
Bibtex Entry:
@misc{van_den_brand_sofie_scripts_2021,
	title = {Scripts for {Post}-{Processing} {Mega}-{Meta} {Screening} {Results}},
	copyright = {MIT License, Open Access},
	url = {https://zenodo.org/record/5752357},
	abstract = {The repository is part of the so-called, Mega-Meta study on reviewing factors contributing to substance use, anxiety, and depressive disorders. The study protocol has been pre-registered at Prospero. The procedure for obtaining the search terms, the exact search query, and selecting key papers by expert consensus can be found on the Open Science Framework. The three datasets, one for each disorder, used for screening in ASReview and the partly labeled output datasets can be found on DANS[NEEDS LINK]. The current repository contains the post-processing scripts to: (1) Merge the three output files after screening in ASReview; (2) Obtain missing DOIs; (3) Apply another round of de-duplication (the first round of de-duplication was applied before the screening started); (4) Deal with noisy labels corrected in two rounds of quality checks. The scripts in the current repository result in one single dataset that can be used for future meta-analyses. The dataset itself is available on DANS[NEEDS LINK].},
	urldate = {2022-07-27},
	publisher = {Zenodo},
	author = {van den Brand, Sofie and Hofstee, Laura and Teijema, Jelle and Melnikov, Valentin and Kramer, Bianca and Brouwer, Marlies and Van de Schoot, Rens},
	month = dec,
	year = {2021},
	doi = {10.5281/ZENODO.5752357},
	keywords = {active learning, ASReview, machine learning, systematic review},
}