Research Data Management

Last update: June 2023

Research Data Management Coordinator: Maike Kleemeyer

Given the increasing amount of (digital) research data as well as the endeavor for openness and transparency, professional research data management (RDM) has become both an essential factor in excellent and sustainable research and a cornerstone of good and open scientific practice. The corresponding activities were initially coordinated by a working group set up by the Institute's Board of Directors in 2018 (see Table 1). One major achievement of this working group was the establishment of RDM as a central service unit, located in the Library and Research Information Unit, with a new permanent position for an RDM coordinator. This position was filled in March 2021 and since then has been coordinating RDM & Open Science-related activities, together with a cross-divisional in-house working group. The coordinator was furthermore the main driver of developing an institutional approach to RDM, involving stakeholders across the Institute in relation to questions of superior strategical significance. To support the implementation of necessary RDM-related software, a project employee was hired in July 2022.

As participant management and recruitment are essential aspects of RDM for human sciences, the team responsible for the development and deployment of the Castellum software was also allocated organizationally to the new RDM unit in May 2022. Castellum has become a powerful tool for participant management that is compliant with the General Data Protection Regulation (GDPR). It is actively used in multiple Institutes within the Max Planck Society (including the MPI for Human Development) and interest has recently been expressed by external research institutions (e.g., Universität Hamburg).

With the growing pressure of funding agencies, journals, and research organizations including the German Research Council (DFG) regarding (raw) data and analysis code sharing, it is increasingly important to educate, in particular, young scientists accordingly. If projects including research data and code are well named, structured, and documented right from the start, the effort needed to openly share them at the end is negligible. A major focus of the RDM & Open Science activities was therefore the organization of a workshop series covering relevant topics, for example, a workshop on the Brain Imaging Data Structure (BIDS). Specifically, the RDM team developed a half-day workshop tailored to the Institute's needs that is regularly offered as part of the workshop series and contains information on all important aspects of RDM, such as data organization, documentation, and storage, as well as final publication and related challenges (e.g., anonymization, licenses). A succinct version of the workshop's contents was also compiled into a comprehensive RDM subpage that has been added to the Library and Research Information Unit’s intranet presence.

Another focus was the provision of general advice and support for researchers wishing to publish their research data, in order to increase the number of high-quality, openly shared data sets. In the reporting period, two large data sets were systematically organized and documented to make them openly available in the near future: In close collaboration with the Leibniz Institute for Psychology (ZPID), data from the Berlin Aging Study (BASE) will be available on ZPID's Research Data Center, and data from the AKTIV study  are about to be published on E-Brains. Provision of access to data from the COGITO Study was transferred to the RDM unit as a first step toward managing access to open data centrally.

Overview of the COGITO Study – www.mpib-berlin.mpg.de/cogito

In the COGITO study, 101 younger adults (20–31 years of age) and 103 older adults (65–80 years of age) participated in 100 daily sessions in which they worked on cognitive tasks measuring perceptual speed, episodic memory, and working memory, as well as various self-report measures (see Schmiedek et al., 2010, 2020). All participants completed pretests and posttests with baseline measures of cognitive abilities and transfer tasks for the practiced abilities. Brain-related measures were taken from subsamples of the group, including structural magnetic resonance imaging (MRI), functional MRI, and electroencephalographic (EEG) recordings. A central goal of the COGITO study was the comparison of between-person and within-person structures of cognitive abilities. Further, the COGITO study qualifies as a cognitive training study of unusually high dosage and long duration because of its 100 sessions of challenging cognitive tasks.

A particular challenge of managing research data at the MPI for Human Development is the multidisciplinarity of research that requires handling highly diverse data from a range of study designs and instruments, which a rather generic workshop cannot possibly cover. The RDM team thus also developed a software tool in cooperation with all Research Centers, conceptually building upon an existing software tool from the Center for Adaptive Rationality. With this study registration tool, a set of meta-information on studies acquiring data is collected, which will allow the provision of customized RDM support throughout the data life cycle. The tool furthermore gives an overview of studies conducted at the Institute, including badges for openness (referring to preregistration, open data, and open materials) to promote scientific exchange between researchers and also the reuse of existing data, especially between researchers from different Centers and Research Groups. The tool was successfully launched in April 2023.

Another important goal of the RDM unit was a more efficient and sustainable way of managing storage space. Together with the Central IT Unit, we linked storage on institutional servers with the new registration tool such that a study folder including predefined subfolders is created when a study is entered. This harmonizes storage across Centers/groups and allows us to closely monitor the allocation of storage on a rather fine-grained level. Furthermore, a procedure was established to ensure responsible data handling when researchers leave the Institute and at the same time adhering to the Max Planck Society's rules of conduct for good scientific practice. Similarly, another aim was to avoid, whenever possible, the arrival of raw DICOM data from the MR scanner on institutional servers. Given that BIDS provides a standardized way of organizing MR data, we developed a systematic workflow that takes the DICOM data from the scanner's storage, directly converts them into BIDS format, and only then stores them on institutional servers. The workflow has been implemented in close collaboration with the Research IT and is now in its final testing phase.

The new RDM services were presented in various contexts, including the 46^th Annual Meeting of the Max Planck Libraries (2022) and the discussion series Human Research Data in Practice organized by the Max Planck Digital Library in 2021/22, and BiblioCon 2023, the most important conference for information professionals from German-speaking countries.