Research Data Management

What is Metadata?

Metadata is, simply, data about data. It is how data is described so that it can be discovered and understood by others. FAIR data is well described data!

How is Metadata Created?

To create metadata you will need to consider how to best describe your data and your research.

The following series of questions from the USGS can be used as a guide through the process of creating metadata:

Metadata in plain language

There are many standards that are used to describe data from various disciplines, as can be seen in the chart below

.

   Metadata Concept Map by Amanda Tarbet is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License.

Metadata standards

There are different standards for metadata that vary by research discipline. They cover such things as minimum description needed, what should be included in the metadata, and recommendations for controlled vocabularies. The link below by the DDC allows you to explore these conventions according to discipline:

Metadata Standards by discipline

Establishing and consistently using a system for naming files can help organize your data and ensure it is understandable to others.

When naming files, here are some things to consider:

• Have you recorded instructions for your file naming scheme (e.g. in a ReadMe file) so that others can follow  and understand it? 
• Version control: how will changes to files be identifiable and tracked?
• Are there conventions used in your discipline for naming or describing things such as languages or dates, such as ISO standards?

Additional information

This guide from Simmons University Library covers naming and organizing files in detail:

File Organization

A pdf file naming guide from Doug Brigham and Eugene Barsky at UBC:

File Naming Guidelines

A short article on file naming best practices by Alistair Downie:

Bite-sized RDM: File naming conventions 

Read me files can help researchers and future users understand datasets. These should be brief, plain text documents that include important information such as the name and dates of the project, contact information, file naming conventions, a description of the data, any restrictions on using the data, and information on how to cite it. Any data deposited into a repository should include a ReadMe file.

Cornell University has an excellent guide for preparing Read Me files, including a template:

Cornell guide to writing Read Me files

Harvard University also has a comprehensive guide, including a video tutorial: 

Harvard guide to ReadMe files

Data can be licensed or copyrighted to establish how it should be used, shared, and to ensure that it is attributed to you.

Creative Commons licenses are a free and commonly used way to license data. You can select the type of license that will best suit your needs:

Creative Commons: Share Your Work

Not sure what creative commons license you need? Try their license chooser tool:

CC chooser

Open Data licenses are another licensing option that was intended specifically for datasets and databases:

Open Data Commons

The article below by Alex Ball for the Digital Curation Centre has more information about data licensing:

How to License Research Data

It's important to identify your data with a permanent identifier like a DOI, but have you identified yourself?

Setting up an ORCID id can help you share your research and ensure your data and articles are attributed to you. It is also required by many major journal publications.

Setting up an ORCID id is free and easy to do. Just click on the link below and follow the steps:

Create your ORCID id

For more information on ORCID ids, please view the PDF below from McMaster University's Sherman Centre for Digital Scholarship, or view the slides at Brief introduction to RDM and ORCID. 

This work was licensed under a Creative Commons Attribution 4.0 International License with attribution due to McMaster University's Sherman Centre for Digital Scholarship.

Often, there are multiple contributors to research data. If you are unsure how to acknowledge them, CRediT (Contributor Roles Taxonomy) is a useful tool that describes various roles. Find it at the link below:

CRediT