UMass Amherst
Site Search
UMass Amherst LibrariesAsk a librarian

Data Management Plan Guidance

Types of Data

Metadata and Standards

Sharing and Access

Re-use and Re-distribution

Short and Long Term Storage and Management

Services » Services for Faculty » Data Management » Data Management Plan Guidance » Metadata and Standards

Metadata and Standards

Metadata describes important information about a data set, such as its content and quality, information about how the files were created and stored, and intellectual property rights and access conditions. Appropriate metadata allows researchers to find, use, and cite your data.

Standards are formal expressions of metadata elements, specifiying what information is appropriate for certain kinds of data and how it should be documented. Some standards are generic, such as Dublin Core (common in bibliographic settings); others are relevant to specific disciplines, such as Darwin Core (for biological sciences), TEI (for textual materials), DDI (for social sciences).

See, UMass Amherst Libraries' Introduction to Metadata.

Things to think about

At a minimum, you should provide enough information about a data set so that it can be cited, such as: 

Creator The main producer of the data or authors of publication where the data can be found.
Title The name by which the data set is known.
Publisher The holder of the data.
Publication Year The year the data was made available.
Identifier The unique identifier for the data.

 

 

 

 

 

For more information about Data Citation, see DataCite.

To provide enough information to make your data set discoverable and useful, MIT Libraries' has general recommendations for data sets of any discipline, see below. Consider all the information that you would require in order to correctly identify, interpret, or reuse the data set and document that information.

Title The title of the data set.
Creator Name of the person, people, or center that created the data set.
Identifier A unique number used to identify the data set.
Subject A description of or keywords describing the content of the data set.
Dates Key dates associated with the data (ie: date recorded).
More....  

 

 

 

 

 

Metadata for Data Sets

A simple way to record metadata for a data set is to create a README.txt file that includes all the metadata that you need to correctly identify and interpret the file.

These examples of structured metadata from the Australian National Data Service are rendered as html files:

Example Language

  1. All electronic data is stored in EXCEL format with each data set including a data dictionary with a description of each data element, the appropriate format, the type of data element, and its intended use.
    (From UCSD to Engineering Directorate, Division of Civil Mechanical and Manufacturing Innovation)
  2. For preservation and long-term access, data collection will be accompanied with proper documentation and associated metadata. Files will include the data itself saved as Matlab (.mat) files, documentation files which a description of how the data was collected, and metadata in the form of the Matlab routines (Psychophysics toolbox) that administer each behavioral task.
    (From UCSD Cross-Directorate)
  3. As data are generated they will be entered into Morpho, a free resource for associating Ecological Metadata Language (EML) with archived datasets. It will be the responsibility of each researcher to annotate their data with metadata, and it will be the responsibility of the PIs to check weekly (during the field season, monthly otherwise).
    (From UCSD Directorate for Biological Science, Division of Environmental Biology)

References

 

 

 

Last Edited: 17 July 2013