Open Science Indicators for a corpus of 8,131 research articles published by Taylor & Francis journals

This dataset represents a set of Open Science Indicators generated by the AI solution provider DataSeer using a corpus of 8,131 research articles published in Taylor & Francis journals in 2023.The corpus was selected using purposive sampling of journal titles to ensure inclusion of Open Access and Open Select (hybrid) journals; journals with a variety of data sharing policies; and journals representing a range of disciplines including life sciences, medicine and health, earth sciences, social sciences and psychology. From each journal in the corpus a random representative sample of between approximately 10-50 articles was selected and the full 2023 published output of any single journal is not included.The DataSeer analysis identified Open Science indicators including:- Presence of data availability statements;- Evidence of data sharing (via supplementary files or data repositories);- Evidence of code sharing;- Evidence of preprinting;- Pre-registration of studies;- Use of persistent identifiers (ORCIDs and RRIDs).As the dataset was generated by an AI tool, some errors or inaccuracies may be present. Before sharing the dataset publicly, the project team at Taylor & Francis undertook data cleansing to ensure that the dataset is comprehensible to an external audience and to enhance its reusability. Notes on data cleansing are included in the README file in the dataset spreadsheet, along with explanations of columns headers where needed.