This repository contains the script to produce one single dataset of records (`smid_2019.csv`) that were screened in subsets as part of the systematic review conducted by [Smid et al. (2020)](https://doi.org/10.1080/10705511.2019.1577140). If you want to use this data in your work, please cite the original publication: Sanne C. Smid, Daniel McNeish, Milica Miočević & Rens van de Schoot (2020) *Bayesian Versus Frequentist Estimation for Structural Equation Models in Small Sample Contexts: A Systematic Review, Structural Equation Modeling: A Multidisciplinary Journal*, 27:1, 131-161, DOI: [10.1080/10705511.2019.1577140](https://doi.org/10.1080/10705511.2019.1577140) The review consists of six different searches, of which the raw data (forthcoming) is in: Smid, S. C., McNeish, D., Miocevic, M., & van de Schoot, R. (2019, September 16). *Bayesian versus Frequentist Estimation for Structural Equation Models in Small Sample Contexts: A Systematic Review.* Retrieved from osf.io/7mght The steps taken to preprocess the raw data are described in the current repository. The resulting dataset is used to simulate an AI-assisted screening process: Gerbrich Ferdinands (2020) *AI-Assisted Systematic Reviewing: Selecting Studies to Compare Bayesian Versus Frequentist SEM for Small Sample Sizes, Multivariate Behavioral Research*, DOI: [10.1080/00273171.2020.1853501](https://doi.org/10.1080/00273171.2020.1853501) # Processing the raw data For every search, TIAB screening was mainly done in JabRef. Full-text screening was done in Zotero. These two screening phases need to be combined to arrive at a set of all records that have been screened at TIAB level, and where the records that were included at full-text level are labeled as such. Preprocessing was done with RStudio. Zotero exports of the TIAB screening are located in the `output/annotated_full-text_screening` directory. The JabRef exports are in the `output/tiab_screening` directory. For searches with missing abstracts, additional abstracts were retrieved via lens, see `output/lens`. ## Search 1 Abstract screening - zotero Full-text screening - zotero Exported from zotero as `output/full-text_screening_annotated/search1.csv`. Process into `fulls1.csv` (script in `search1.R`) ## Search 2 Abstract screening in “Search2Jabref” - Open "Search2Jabref.sav" in JabRef - Save as `output/tiab_screening/search2.bib` (2173 references) Fulltext screening zotero “Search 2 – 167 ref" - 167 abstract inclusions, 8 final - added manual tag 'Included' for the 8 final inclusions - Save as `output/full-text_screening_annotated/Search 2 - 167 ref.csv` Combining both files - `output/complete_search/fulls2.csv` (script in `search2.R`) Retrieve extra abstracts - result in `output/complete_search_with_additional_abstracts/fulls2abstr.csv` (script in `get_abstracts_s2.R`) ## Search 3 Abstract screening in "Search 3 – mailing lists" Fulltext screening in "Search 3 – mailing lists" Exported as `output/full-text_screening_annotated/search3.csv`. Process into `output/complete_search/fulls3.csv` (script in `search3.R`) ## Search 4 Abstract screening in jabref “Search4Jabref” (export in `output/tiab_screening/search4.bib` ) Full-text screening via Zotero in folder genaamd “Search 4”, export `output/full-text_screening_annotated/search4.csv`. Combining both files - `output/complete_search/fulls4.csv` (script in `search4.R`) Retrieve extra abstracts - `output/complete_search_with_additional_abstracts/fulls4abstr.csv` (script in `get_abstracts_s4.R`) ## Search 5 Abstract screening in jabref called “Search5Jabref", output in `output/tiab_screening/search5.bib`. Full-text screening via zotero in “Search 5”, output in `Search 5_abstract_included.csv`. Combining both files - `output/complete_search/fulls5.csv` (script in `search5.R`) Retrieve extra abstracts - `output/complete_search_with_additional_abstracts/fulls5abstr.csv` (script in `get_abstracts_s5.R`) ## Search 6 Screening via excel, `output/full-text_screening_annotated/Search6.xlsx` Process and detect abstract and full-text inclusions in `output/complete_search/fulls6.csv` (script in `search6.R`). # Creating the final dataset The script `combine_all_searches.R` takes all processed searches and combines them into 1 final dataset, `smid_2019.csv`. This data set can be used for the simulation mode in ASReview. # License This project has a CC-By 4.0 license.