space-walk-991_960_720.jpg

Faculty of Aerospace Engineering

Data Steward and author: Heather Andrews

Overview

The Data Steward at the Faculty of Aerospace Engineering (AE), Heather Andrews, started working in October 2017. So far her main duties have been:

  • undergoing training in research data management;
  • performing qualitative interviews of staff members to understand the current data management practices and issues within the different research groups in the Faculty;
  • working on strategies to introduce the Data Stewardship project to the staff members and create awareness about Open Science and responsible data management;
  • monitoring and promoting the Research Data Management (RDM) survey that has been launched in each Faculty that has a Data Steward.

In the following sections we will cover:

  • Regular meetings
  • Training sessions
  • Qualitative interviews with researchers
  • RDM survey results
  • Scheduled appointments
  • Plans for 2018

Regular Meetings

The Data Steward has weekly meetings together with the other Data Stewards and the Data Stewardship Coordinator, in order to exchange practice about data stewardship work at Faculty levels, and discuss matters regarding the Data Stewardship Project at a university level (1.5 hours weekly).

In addition to that, the Data Steward meets the secretary of the AE Faculty every 2 weeks to discuss progress within the Faculty itself (1 hour every two weeks).

The Data Steward must also attend the meetings of the Library Committee of the AE Faculty, which are once every 3 months.

Aside from these regular meetings, other appointments were carried out in the first weeks, in order to get to know important support staff members of the Faculty and to have an optimal overview of how the AE Faculty works.

Training Sessions

The Data Steward is undergoing two different types of training on research data management: internal and one external. The main goal of these training sessions is to get a broad overview of data management topics and learn the essential skills to support researchers with their data questions.

As part of the internal training, the Data Steward has attended 2 training sessions delivered by local experts and organised specially for the Data Stewards at TU Delft: one on Data Management Plans, and the other on central ICT solutions for researchers (2 hours each).

As part of the external training, the Data Steward has been enrolled in the Essentials 4 Data Support course organised by SURFSara, 4TU.Centre for Research Data and DANS. This course consisted of 16 hours of in-person training: on October the 5th and on November the 23rd, plus 4 hours a week (on average) of doing the assignments and studying the provided online material.

Aside these training sessions, the Data Steward has also spent time self-training, as she arrived 1.5 months later than the other Data Stewards.

Qualitative Interviews with Researchers

During this 1.5 months period, the Data Steward has carried out several interviews with researchers within the AE Faculty (from all four departments), in order to get to know the data management practices within the different research groups. Up to 16-11-2017, 13 interviews were completed (1 Postdoc, 1 Associate Professor, 8 Assistant Professors and 3 Professors). Among these, 3 researchers are from the Space systems Engineering (SpE) department; 2 are from Control & Operations (C&O); 4 are from Aerospace Structures and Materials (ASM); and 4 are from the Aerodynamics, Wind Energy, Flight performance & Propulsion (AWEP) department. There are 3 more interviews scheduled in the upcoming weeks (with Professors only).

Main issues discussed during the interviews are about data storage, archiving, and sharing. From talking to the researchers, it is clear that every research group is different and has specific needs regarding data management. For example, some groups need to continuously have access to 10-year-old data, while others might only require access to research data from the past year. There are groups that generate so much data they have to process it in Supercomputers, and reduce it to sizes which could be retrieved relatively fast to work on their local disks. There are groups which keep all their data on their own external hard drives, which costs more than 10-15k euros a year.

Overall trends can be summarized as follows

  1. There is an important lack of information regarding services provided by TU Delft. Either researchers have never heard of available services, or they have heard something but have not spent the necessary time on getting to know more about them.
  2. All interviewees are willing to share their data once an article is published (and as long as it is allowed when working with companies).Regarding sharing unpublished data, most researchers are open to the idea. However, the main problem is finding the time to put such datasets in a  comprehensible format, which would be useful for the community. Most researchers claim that this subject is not viewed as a ‘high priority’. Unless researchers see a direct benefit, they will invest time in making all their data “human-readable”. Therefore, most researchers currently share data under private requests, meaning that whoever is interested in the data, sends an email explaining why they want the data, and then the researcher evaluates whether it is pertinent to share the data or not.
  3. Some groups are now starting to have standard procedures regarding data storage. This has been motivated by the fact that whenever someone leaves the group, the data is either lost or delivered in a non-understandable format.
  4. There are groups interested in using a repository to save all their data during research.
  5. SURFdrive is not always the preferred choice for data sharing. Some staff members claim there are technical problems with SURFdrive and it is not very user-friendly when it comes to accessing the data from abroad or sharing the data with colleagues outside of academia.
  6. Some groups save all their research data files on external hard drives.
  7. Researchers would appreciate more awareness and training about good data management practices, which was also confirmed in responses to the quantitative RDM survey (see below).

Research Data Management Survey

Preliminary results of the RDM survey conducted at the AE Faculty (answers gathered between 26/10/2017 and 14/11/2017) can be summarized as follows:

  • More than 27% of the staff members who received the survey, filled it in.
  • Among the people who replied to the survey, 3% are MSc students, ~48% are PhDs, 14% are Postdocs/Researchers, 17% are Assistant Professors, 10% are Associate Professors and 8% are Full Professors. Considering the total number of PhD students who received the email, we have about 29% response rate to the survey. Estimate response rate for Full Professors is 52%.
  • Most of the people who replied to the survey (78%) are unaware or not sure of what FAIR data is.
  • 67% of the people who replied to the survey have not heard about the Data Stewardship project and the dedicated support for data management at the Faculty (9% of the surveyed people are ‘not sure’ whether they have heard of them).
  • 53% of the people who replied to the survey claim to know who owns the data they work on, while 33% do not know who owns it and the rest simply do not know.
  • 44% of the PhDs that replied to the survey claim to know who owns their data. Out of these, 16% of them claims to have full or partial ownership of the data. In case of partial ownership, PhDs think that the data ownership is shared between them and either TU Delft or their promoter.
  • Most of the people who replied to the survey do not have data management plans for their research projects (63%), or are not sure of having one (20%).
  • 35% of the people who replied have their data automatically backed up.
  • 85% of the surveyed people are interested in training on research data management. Only 17% of the PhDs and 14% of the Postdocs/Researchers that replied to the survey claimed to be ‘Not interested in any training’.
  • 51% of the people who replied to the survey are not aware of the 4TU.Centre for Research Data.

A more in-depth analysis will be performed once the survey is closed. It is worth mentioning that also the interviewees mentioned in Section 3 have all expressed great interest in knowing what comes out of this survey.

Plans for 2018

During the first 3 months of 2018, the plan is to make sure the Data Stewardship project has been properly introduced to the community at the AE Faculty. In addition, the results of the RDM survey will be presented to the staff members, to create awareness of the current data management practices across the Faculty, and also to show how they compare to the practices in the other 2 Faculties that were surveyed. From the Data Stewardship project perspective, the results of this survey will also be used to benchmark the impact of the project.

The Data Stewardship project will set up new guidelines and policies regarding research data management; a work that has already started this year but it is expected to be implemented from 2018 onwards.

The Data Champions program will start in 2018. This program aims to create a community of researchers who are actively engaged with responsible data management practices. The Data Champions will work at a department-level within the Faculty, helping with the implementation of discipline-specific policies, being coordinated by the Data Steward. First, the program will be presented to researchers, to look for interested candidates. Candidates will be assessed at the beginning of 2018, so that the selected ones (the ‘Data Champions’) can start working starting from March-April 2018.

Given that the training sessions described in Section 1 will be completed by the end of 2017, it is expected that throughout the year 2018 the Data Steward can start actively helping researchers regarding their data management issues; focusing on problems that were mentioned by researchers during qualitative interviews, but also on new projects across the Faculty to make sure that their data management is well set from the beginning.

Aside from providing data management support, the Data Steward will also focus on creating awareness about Open Science among researchers and will provide training for the community (e.g. to PhD students) on responsible research data management practices and data management services provided by the TU Delft (e.g., 4TU.Centre for Research Data).

Finally, at the end of 2018, the RDM survey will be ran again to see what the differences are with respect to the beginning of the Data Stewardship Project.