Dataset for an Open Science Dashboard: Open Research Data and Data Reuse in Earth Sciences at the Department of Earth Sciences, Freie Universität Berlin

Openly shared and reused research data are increasingly recognized as important indicators of transparent and reproducible research practices. However, monitoring the prevalence of open data in the scholarly literature remains methodologically challenging due to the lack of standardized, discipline-specific detection workflows. This dataset provides the results of such a workflow, developed to detect open and reused research data underlying journal articles published by researchers affiliated with the Department of Earth Sciences at Freie Universität Berlin (FU Berlin) between 2016 and 2024. The results are visualised in the Open Science Dashboard for the Department of Earth Sciences, Freie Universität Berlin in tab Open Data. The underlying publication list (n = 2.186 articles) was obtained from the university library's bibliography. Open data mentions were detected using an Earth Sciences-adapted version of the ODDPub text-mining algorithm (Riedel et al., 2020), following a workflow adapted from an existing biomedical detection approach (Iarkaeva et al., 2024). The algorithm's list of field-specific data repositories was extended for the Earth Sciences using repository names queried from the re3data.org (Registry of Research Data Repositories). Articles flagged as potentially mentioning open datasets by the algorithm were manually validated using the Numbat screening tool, applying a set of verifiable open data criteria (Bobrov et al., 2024) covering data findability, persistent accessibility, authorship, machine-readability, and rawness. The dataset comprises following files: Data dictionary (data_dictionary_open_data_earth_sciences_fu_v1.csv) — description of all columns and values in the data files. Article-level file for all journal articles (open_data_earth_sciences_fu_all_journal_articles_processed_2016_2024_v1.csv) — complete list of 2.186 journal articles published between 2016 and 2024, indicating for each article whether it was screened, why articles were excluded from screening (e.g. PDF not retrievable), and the final validated open and/or reused data status. Dataset-level file for only Open Data positive cases (by ODDPub flagging)(open_data_earth_sciences_fu_dataset_level_processed_2016_2024_v1.csv) — 392 manually validated dataset-level extractions from 312 articles flagged as open-data-positive, including repository names and types (disciplinary vs. general-purpose), data identifiers, and outcomes of the manual validation criteria. Article-level file for only Open Data positive cases (by ODDPub flagging)(open_data_earth_sciences_fu_article_level_processed_2016_2024_v1.csv) — article-level extractions for 312 journal articles flagged as open-data-positive, including the outcome regarding open, reused, and restricted data at the article level (i.e., whether an article had at least one openly shared, reused, or restricted dataset), and repository type (disciplinary vs. general-purpose). An overview of the main results is presented in the following table: Level File Column TRUE (N) FALSE (N) Not accessible (see reasons in data_dictionary_od_earth_sciences_fu.csv) All journals (2) open_data_result_processed 154 158 1874 All journals (2) pdf_accessed_processed 1722 417 46 All journals (2) reuse_data_result_processed 84 228 1874 All journals (2) restricted_data_result_processed 4 308 1874 All journals (2) open_data_checked_manually 312 1411 463 This dataset may be reused to study research data sharing and reuse practices in the Earth Sciences, to compare disciplinary patterns of open data adoption with other fields or institutions, or to inform the further development and validation of automated open-data-detection workflows in other disciplinary contexts.--------------------------------------------------------------------------------- Bobrov E, Riedel N, Kip M (2024) Operationalizing open and restricted-access data—Formulating verifiable criteria for the openness of data sets mentioned in biomedical research articles. Quantitative Science Studies; 5 (2): 383–407. doi: https://doi.org/10.1162/qss_a_00301 Iarkaeva A, Nachev V, Bobrov E (2024) Workflow for detecting biomedical articles with underlying open and restricted-access datasets. PLoS ONE 19(5): e0302787. https://doi.org/10.1371/journal.pone.0302787 Riedel, N., Kip, M. & Bobrov, E. (2020). ODDPub – a Text-Mining Algorithm to Detect Data Sharing in Biomedical Publications. Data Science Journal, 19(1), 42. https://doi.org/10.5334/dsj-2020-042