Code in Ecology

This repository contains the data, code and other materials used in the following study:

---

Antica Culina*, Ilona van den Berg, Simon Evans, Alfredo Sánchez-Tójar*. 2020. Low availability of code in ecology: a call for urgent action. PLoS Biology, 18(7): e3000763. DOI: 10.1371/journal.pbio.3000763

---

The supplementary information is available at: supplementary information

* For any further information about the code, please contact: Alfredo Sánchez-Tójar, email: alfredo.tojar@gmail.com.

* For any further information about the data, materials or more generally, the study, please contact Antica Culina, email: A.Culina@nioo.knaw.nl; and/or Alfredo Sánchez-Tójar, email: alfredo.tojar@gmail.com.

Information about scripts, folders and files within:

data: contains the data used in the study. Specifically, it contains five files:

• Code_policies.docx: word document listing the journals with code-sharing policies in March 2020 (more in supplementary information) and a copy-paste of the information provided in the author guidelines/journal policies section of each journal.
• Data_Feb_2020_V8.xlsx: excel document with three sheets containing the data extracted during the review process, including the data regarding those articles that were double-checked by a second observer: (i) Dataset_2015_16: data for the 2015-2016 period (extracted by AC, double-checked by AST); (ii) Dataset_2018_19: data for the 2018-2019 period (extracted by AST, double-checked by AC); and (iii) explanations: metadata provided information about the variables extracted in the first two sheets.
• Updated_Table_Mislan_2020_v2.xlsx: excel sheet containing data about the type of code-sharing policies of each of the 96 journals reviewed in March 2020. The list of journals is based on that compiled by Mislan et al. 2016.
• code_availability_full_and_clean.csv: this is the final dataset used for the analyses conducted in this study, and this dataset is the output of the script 003_data_cleaning_and_standardization.R, which imports the file Updated_Table_Mislan_2020_v2.xlsx, cleans it, and creates new variables for the analyses.
• journal_percentages.csv: small dataset created by the script 005_supporting_information.Rmd and containing each of the 14 journals reviewed, the number of articles reviewed, the number of articles sharing at least some code, and the percentage of articles sharing code for each journal. This dataset is used to create figure 2, which was originally included in the supplementary information but now available in the main text.
• journal_info_v2.csv: small dataset containing the code-sharing policy and name abbrevation of each of the 14 journals from which we calculated code-sharing practices. Note that the code-sharing policy of this journals is also available in the file Updated_Table_Mislan_2020_v2.xlsx.

docs: contains the supplementary information file of this study. To visualize the html file, click here.

plots: contains all four figures from this study. Figure 1, 2 and 3 correspond to those included in the main manuscript, and Figure S1 is included in the supplementary information.

screening_process: contains data for reproducing the literature search of 2018-2019. For the literature search of 2015-2016, no code was used (except the function sample() for ramdonly selecting articles, which was used for both years (code not available). This folder contains two subfolders:

1. title-and-abstract_screening: containing multiple files used for the title-and-abstract screening performed in Rayyan.

• random_200_2018_2019.bib: all 200 references reviewed for 2018-2019 in .bib format.
• random_200_2018_2019_rayyan.csv: all 200 references reviewed for 2018-2019 in .csv format, with only the necessary information (variables) for importing into the title-and-abstract screening software (Rayyan). This file is the output of the script 001_title-and-abstract_screening.R.
• random_200_2018_2019_rayyan_edited.csv: same file as random_200_2018_2019_rayyan.csv after removing some quotation marks manually so that it can be imported into the title-and-abstract screening software (Rayyan).
• rayyan_csv_example.csv: template with the variables needed for importing into the the title-and-abstract screening software (Rayyan).
• screening_process_Rpackages_session.txt: information about the R session (e.g. versions and packages used) used when running the script 001_title-and-abstract_screening.R.

2. fulltext_screening: containing multiple files used for the fulltext screening:

• instructions/Code_evaluation_template_v2.xlsx: excel file used as template for choosing what variables to add.
• fulltext_screening_process_Rpackages_session.txt: information about the R session (e.g. versions and packages used) used when running the script 002_fulltext_screening.R.
• random_200_2018_2019_rayyan_edited_t-and-a_decisions.csv: .csv file extracted from Rayyan and containing the title-and-abstract decisions taken during the screening.

Culina_et_al.bib: bibliography used in the supplementary information file only.

Scripts

001_title-and-abstract_screening.R: R script used to format a reference dataset (.bib) so that it can be imported into Rayyan for title-and-abstract screening.

002_fulltext_screening.R: R script used to subset a reference dataset after title-and-abstract screening was performed in using Rayyan, and prepare the fulltext screening datasets.

003_data_cleaning_and_standardization.R: R script used to import and clean the sheets with data extracted from the reviewed articles, and generate a final and clean data set (i.e. code_availability_full_and_clean.csv).

004_plotting.R: R script used to generate Figure 1 and 2 from the study.

005_supporting_information.Rmd: Rmarkdown script used to generate the supplementary information file of this study (see here).

Notes:

See details of the licence of this repository in LICENSE.txt