SEA-SEARCH = Guidelines and Practices / MAST: A Guideline for Project Data Management

MAST: A Guideline for Project Data Management
Mast Data Committee - March 1997, MAST Data Committee

This guideline as well as the "Guideline for Better Practice in Data Documentation" is written to help teams to implement the "Code on Data Management in MAST Projects".

1. INTRODUCTION

As each project differs in terms of scale, scientific rationale and short and long-term objectives, it is not appropriate to devise a generic data management plan. Specific plans must be designed to suit each project, taking into account the type and volume of data, as well as technical and financial constraints. Thus, this guideline gives a summary of the overall objectives of a data management plan and outlines basic considerations which should be addressed when preparing a data management solution.

The objectives of project data management are:

To maximise the consistency of data collected throughout the project by ensuring that data are documented to an agreed standard and that reliable information is given to assess the quality and limitations of the data and therefore it's suitability for a particular task.
To facilitate data and information exchange within the project and thus maximise the use of the data during the course of the project.
To ensure the longevity and long-term multiple re-use of the data as a coherent data set through publication.

2. PREPARATION OF THE DATA MANAGEMENT PLAN

It is recommended that the data management plan be formulated as part of the preparation of the scientific proposal. An assessment of the project "data requirements" should be made at an early stage. This facilitates the preparation of the scientific proposal and of a realistic data management plan.

Considerations should be given to existing solutions, approaches, standards etc. and facilities that may be cost-effectively used or adapted for use in the project. For most projects, data management is part of the routine core activities of the project and not part of its development front.

The data management plan should comprise of a series of well constructed, appropriate measurable deliverables to support the scientific plan.

The partners responsible for the data management should give a clear picture of their ability to achieve these deliverables.

It is also recommended that at an early stage, a "Data Advisory Team" is appointed which comprises of a selection of scientists and data managers. The role of the selected scientists will be to advise on data management issues concerning their particular scientific discipline or research area.

3. CONTENT OF THE DATA MANAGEMENT PLAN

The data management plan should at the very least, document how the data management partner and science partners will address the following points:

INFORMATION AND DATA FLOW: The data management plan should show how it will assist and promote the flow of data and information within the project.

DATA DOCUMENTATION: The data management plan should describe how data will be documented so that sufficient information will be included to accurately describe all data and to allow an assessment of the data's quality and limitations.

DATA QUALITY: It is the ultimate responsibility of the data producer to ensure the quality and integrity of the data. However, there are a number of ways in which the data management plan can address this issue. The plan should ensure that quality information is collected for each data set so all processes used to ensure data quality are identifiable. The data management team or external experts may also play a quality checking role on data in many cases.

TECHNICAL ISSUES: Any database applications developed for the project data should ensure easy retrieval and distribution of the data.

DISSEMINATION: To ensure that the data is maintained for future multiple reuse as well as present use, the final project data should be disseminated as a coherent data set through publication.

LONGEVITY AND FINAL ARCHIVING: To maximise the multiple use of the data and its longevity, the resulting published data set(s) should be banked at a suitable institute for long-term archival. At the project proposal stage the archival institute must agree in writing to accept the data for its long-term archive.

MEETINGS: In order to be successfull, data management works in parallel with the scientific progress and requires very close cooperation of all participating individuals. Meetings between data managers and selected scientists (e.g. Data Advisory Team) are essential to design, implement and run a good data management solution. Depending on the size of the project, the number of meetings required may vary. As a guideline, it is recommended that where possible, a "data management" meeting is organised at the proposal stage, again when the project has commenced, and during the course of the project at any major event (main cruise, annual meeting, test run of new laboratory facility, etc.).

PERFORMANCE INDICATORS: The data management plan should identify a series of milestones which will act as performance indicators for the successful implementation of the data management plan.