Marine Biology Undergraduate Thesis Corpus of the Universidad de Magallanes (2008–2025): Metadata, Taxonomic Records, Research Themes and Scientific Outputs

This repository contains the analytical data and reproducibility materials supporting the study "Undergraduate thesis repositories as records of scientific knowledge production and scholarly communication: analysis of a marine science corpus from 2008 to 2025". The study examines the complete corpus of 80 undergraduate Marine Biology theses produced at the Universidad de Magallanes, Chile, between 2008 and 2025. The revised deposit includes the curated thesis-level analytical dataset, controlled vocabularies and standardized descriptors used in the content analysis, derived and curated page-level analytical data, documentation and Python code describing the computer-assisted text-mining workflow, and relational datasets linking undergraduate theses with subsequent publications and conference presentations. The full-text analysis was conducted on a page-by-page basis using controlled term matching followed by manual contextual review. The page-level analytical corpus comprised 5,710 pages. Eight thesis documents required optical character recognition (OCR). Taxonomic nomenclature was standardized using the World Register of Marine Species (WoRMS). Stable Thesis_ID identifiers are used where personal identification is not required. Names are retained only when they form part of the original public scholarly record. The original thesis PDF files are not included because they remain publicly accessible through the Marine Biology Program repository of the Universidad de Magallanes. This revised version expands the original Zenodo deposit to provide additional analytical documentation, intermediate datasets, controlled vocabularies, and reproducibility materials in response to peer-review recommendations. File Description README_analytical_workflow-theses-bm-umag2008-2025.txt Description of the thesis corpus, document preparation, OCR, analytical workflow, manual curation, analytical units, linkage procedures, and limitations. text_mining_workflow-theses-bm-umag2008-2025.py Python implementation documenting the controlled term-matching, page-level counting, and normalization workflow. controlled_vocabulary-theses-bm-umag2008-2025.xlsx Controlled terms, variants, synonyms, standardized descriptors, analytical domains, and explanatory notes used in the content analysis. derived_page_level_data-theses-bm-umag2008-2025.xlsx Derived page-level textual-prevalence values used in the taxonomic, environmental, methodological, and funding analyses. curated_page_level_analysis-theses-bm-umag2008-2025.xlsx Curated page-level presence matrix and analytical summary tables underlying the derived page-level results. thesis_level_dataset-theses-bm-umag2008-2025.xlsx Curated thesis-level analytical dataset using stable Thesis_ID identifiers. Table_S2_thesis_publication_links-theses-bm-umag2008-2025.xlsx Relational dataset linking undergraduate theses with subsequent publications and documenting the basis of each relationship. Table_S3_thesis_conference_links-theses-bm-umag2008-2025.xlsx Relational dataset linking undergraduate theses with conference presentations, distinguishing 49 unique presentations from 50 thesis–presentation associations.