Classifying Data Management Plan Guidance using DMPTool

Objectives: To understand the kinds of guidance R1 universities give researchers for the development of data management plans by analyzing the themed guidance text they provide in DMPTool.

Methods: DMPTool is a popular open-source online application for writing data management plans using generic and funder-specific templates. Institutions can become members of DMPTool, allowing them to provide their students, faculty, and staff with an institutional login, the ability to request feedback on their plans, and access to institution-specific guidance. 92% of R1 universities are members of DMPTool and 48% provide guidance for at least one of the 14 data management themes available in DMPTool, such as data format or preservation. This makes DMPTool a convenient source to examine the data management guidance research-intensive American universities are providing their constituents, which would otherwise likely be limited to their respective websites. For this study, themed guidance text provided by 70 R1 universities was copied from DMPTool into Excel. Each link in the text blocks was then coded according to which of nine destination types it pointed to, such as a library website, university research website, or an outside resource. The average text length and number of links were compiled, as well as the most popular guidance themes and the most common linked resources.

Results: The analysis of 490 themed guidance text blocks from 70 R1 institutions showed that guidance text length ranged from 2 words to 2,259 words, and the largest number of links provided in one text block was 31. Universities provided guidance on an average of 6.93 out of the 14 possible themes. Data sharing was the most common theme, while data collection was the least common. The three most common link destinations were the university’s library website, an independent outside resource, and a university research website, and the three most shared outside resources were re3data, the journal Nature, and DataOne.

Conclusions: This study found a wide variety of approaches to advising researchers on writing data management plans, from minimal to detailed guidance. Although best practices for research data management exist, there is no standard for how to deliver this guidance to researchers, especially in an interactive environment like DMPTool which presents unique challenges for information delivery. The assembled text corpus offers opportunities for further research utilizing topic analysis to evaluate the text on a more granular level. Additionally, this dataset could be compared to a second snapshot taken at a later date to see how data management plan guidance shifts over time or compared to guidance
provided by a different set of institutions.

Jones, L.C. & Phegley, L. (2023, March 16-19). Classifying data management plan guidance using DMPTool [Poster presentation]. 2023 Hybrid Medical Library Association and Special Library Association meeting, Detroit, MI. https://www.mlanet.org/page/mlasla23